See also: Version 6 FAQ

Click here for Version 6 FAQ

Frequently Asked Questions (Version 5.5): Table of Contents

Custom Programs:

User-Exchange Forum/Library of STATISTICA BASIC Programs (Version 5.5) Visual Basic/STATISTICA Interface Programs (Version 5.5) STADEV Example Programs (Version 5.5)

Installation Instructions for STATISTICA Products

Our newsletter: STATSOFT NEWS

Current Issue Back Issues

    User-Exchange Forum/Library of STATISTICA BASIC programs (Version 5.5). A library of example programs (many of which were written by users of STATISTICA) is available (see User-Exchange Forum/Library of STATISTICA BASIC Programs). These programs include a wide variety of advanced illustrations of how to create highly customized reports, graphs, follow-up analyses that use STATISTICA output as input, and how to design your own complete application programs (e.g., statistical procedures) with user interface, and Scrollsheet and graphics output.

    Library of Visual Basic Examples (Version 5.5). A library of example programs which show how to call STATISTICA (including STATISTICA BASIC) is available (see Visual Basic/STATISTICA Interface Programs). These programs include examples of calling STATISTICA from within Visual Basic, calling STATISTICA BASIC from within in Visual Basic and managing STATISTICA files from within Visual Basic via STATISTICA Development Environment.

    Library of STADEV example programs (Version 5.5). A library of STADEV example programs is available (see STADEV Example Programs). These programs include illustrations on using STADEV in a variety of programming environments including Visual C++, Visual BASIC, Delphi, and STATISTICA BASIC.

    On-line documentation (Electronic Manual). The quickest way to obtain comprehensive information on each of the topics mentioned here (including examples and technical notes) is to use the STATISTICA Electronic Manual -- that is, the context-sensitive Electronic Manual accessible by pressing the toolbar Help button, the F1 key, or double-clicking on the status bar (on the bottom of the STATISTICA window). The Electronic Manual offers Hypertext facilities, contains a Glossary, and provides a complete reference to all features of the system.

StatSoft Home Page

TABLE OF CONTENTS

TROUBLESHOOTING PROBLEMS

Troubleshooting problems during installation of STATISTICA

The STATISTICA Concurrent Network Version workstation (i.e., client) installation finishes successfully. But I receive an error message that says "Active-X component cannot create object" when I try to logon to STATISTICA for the first time from this workstation.
   Under certain circumstances, the installation procedure does not successfully install a necessary system update. If you receive this error message, please download a supplementary installation routine called fix_COMDLG32.exe. This additional setup program is to be run, on each machine that experiences the problem, after the Concurrent Network Version workstation installation is completed.

At the end of the installation procedure (before installing ODBC drivers) a General Protection Fault error message is issued and the installation process is aborted. How can I install STATISTICA correctly on my computer?
    This problem is occasionally reported when STATISTICA tries to install ODBC-related dynamic link libraries (DLL) in your WINDOWS/SYSTEM directory where older versions of these files already exist. There are two workaround solutions to this problem:

1. If you don't need ODBC drivers for data import into STATISTICA, choose the option for custom installation during the STATISTICA setup procedure and de-select the ODBC driver option.
2. If you do need ODBC drivers with STATISTICA, look for any the following files in your WINDOWS/SYSTEM folder:
   CPN16UT.DLL, msAJT200.DLL, msJETERR.DLL,
   msJETINT.DLL, msTX2016.DLL, msXL2016.DLL,
   ODBC.DLL, ODBC16UT.DLL, ODBC32.DLL,
   ODBCCP32.DLL, ODBCCURS.DLL, ODBCINST.DLL,
   ODBCJT16.DLL, ODBCTL16.DLL, ODDBSE16.DLL,
   ODEXL16.DLL, ODFOX16.DLL, ODPDX16.DLL,
   ODTEXT16.DLL, PDX200.DLL, VAEN2.OLB,
   VBAJET.DLL, VBAR2.DLL, XBS200.DLL
   If any of those files are found, make a backup copy and delete the file(s) in the WINDOWS/SYSTEM folder. Then start the STATISTICA setup again.

After the installation procedure I get an error message about a DDE communication failure and the setup program is aborted. Does this mean that STATISTICA is not correctly installed on my computer?
    No, STATISTICA is correctly installed on your system. However, we received occasional reports of this non-fatal error message. It usually indicates that the setup program could not create a STATISTICA program group. You can easily create a STATISTICA program group (Windows 3.x) or desktop shortcuts for your STATISTICA modules (Windows 95/NT 4.0) yourself. For details on how to create a program group or shortcuts (desktop icons) please consult your Windows documentation or the STATISTICA Quick Reference.

I tried to run the setup program, but I got the following error message: "Cannot find _mssetup.exe." and the installation failed. What did I do wrong?
    This error message only occurs when installing STATISTICA from 3.5" diskettes. You have most likely accidentally deleted the file _msssetup.exe which is located on your first installation disk by trying to execute it. Since this file is vital for the setup program, you need to obtain a copy of this file and put it onto your first installation disk. If the file does not fit on your first disk, you can move the txt files to another location in order to free up disk space. Note that the file _mssetup.exe is available in the DOWNLOAD area of our web site.

After installing STATISTICA on a system running Windows NT, every time I try to run the program I get a message saying the environment is too large. What is the problem?
    This problem used to occur on some installations of Windows NT 3.xx; however, it has been corrected in programs which were purchased after July 20, 1997. Contact StatSoft at info@statsoft.com for more details.

Troubleshooting STATISTICA Electronic Manual problems

When I try to access the Electronic Manual for STATISTICA, I receive the following message: "Cannot display this help file. Try opening the help file again, and if you still get this message, copy the help file to a different drive, and try again. (136)." How can I access the Electronic Manual?
    According to ms, this behavior can occur if the help file you are trying to open is currently open on a computer running ms Windows NT. It appears to us that the problem is more generic and happens as users try to access shared files. It also seems to be fixed if you delete the STATHELP.GID file which is located in the STATISTICA program directory. The STATHELP.GID file is an index of keywords that Windows generates to optimize access to the help files. Deleting will only cause Windows 95 or NT to create it again.

Troubleshooting general problems running STATISTICA

When I print a 3D graph in STATISTICA the scale labels are missing, even though the graph looks fine on the screen. How can I print the complete graph?
    This problem is occasionally reported with some printers under Windows NT. It usually occurs if your printer driver does not support a 3D representation of fonts properly. In order to print out the 3D graph correctly, de-activate the 3D Text check box in the General Layout dialog for the graph. You may also contact your printer manufacturer or hardware dealer for updated printer driver versions.

When I try to run STATISTICA I receive a message "Cannot read from drive A." If I click on Cancel, STATISTICA opens with the a new empty spreadsheet. What is the problem?
    You probably saved a STATISTICA data file on a diskette the last time you exited the program. STATISTICA usually keeps a log of the recently used files and opens the last file that you saved. Therefore, the program is looking for the last file on the A: drive (or any other removable media drive). However, if your floppy disk drive is currently empty, STATISTICA cannot find it. If you have set the number of entries in the list of recently used files to 1 in the STATISTICA: General Defaults dialog, this message can reoccur everytime you open STATISTICA. In order to prevent it, you can either temporarily set the number of recently used files to 0 or delete the STATIST.INI file (STATISTICA will recreate this file next time you open the program). Note however, that you will lose customized STATISTICA settings if you delete the INI file. Alternatively, you might modify the list of the recently used files in the [file list] section of the STATIST.INI file.

I run Windows 95/NT, but all of sudden my dialogs for opening or saving files in STATISTICA appear in Windows 3.1 fashion. Also, long file names are not supported any longer. How can I make STATISTICA behave like a Windows 95 application again?
    There are two possible solutions to this problem. First check you Statist.ini file for "LongNames=number" This turns off the long file names in STATISTICA. Removing this line from the Statist.ini file and placing Sta_lfn.exe in the STATISTICA directory (C:/STAT by default) will ensure that STATISTICA continues to use long file names.
    If the line described above is not in your Statist.ini file, your registry information is probably damaged. STATISTICA runs on multiple Windows versions including 3.xx, Win95, and NT. If your STATISTICA registry information is damaged, it might not recognize your operating system as Windows 95 and therefore defaults to Windows 3.xx behavior. You need to re-install STATISTICA again in order to have it behave like a Windows 95 application.

Question: When launching STATISTICA under Windows NT, I get a "cannot run XX-bit application" message.

Cause: This problem may occur when a program attempts to access an incorrect version, outdated, missing, or damaged DLL file.

Resolution: Expand the following files from the original Windows NT compact disc to the \System32 directory:

Compobj.dll
Ddeml.dll
Ole2.dll
Ole2disp.dll
Storage.dll
Ctl3dv2.Dll
Ole2nls.dll
Stdole.tlb
Typelib.dll
Ver.dll

NOTE: The Ver.dll file is located in the %SystemRoot%\System folder and the %SystemRoot%\System32 folder, and both versions have the same file size and date.

To complete this file "expansion" from the NT4.0 CD carry out the following steps: [Note: it is very important to make sure you have a good emergency repair disk for your NT 4.0 workstation before you complete the following steps.]

1. From the Windows Start button, select "Run"
   
2. In the file to open box, type in the following command: Expand e:\1386\filename.dl_%SystemRoot%\System32\filename.dll Note: substitute for "e" the drive letter where your NT4.0 CD is located. Substitute the actual name of the file for expansion in the place of "filename". For example: "Compobj.dl_"
   
3. Click . Complete this process for all listed files.
   
4. Reboot your PC.
   
5. Make an emergency repair disk of your NT 4.0 system now: place new diskette into floppy drive and type, "Rdisk".

Troubleshooting problems with data import

Why do I get an error message when trying to import a Paradox (.db) file?
    The STATISTICA Quick Import facility does not support files from Paradox, release 5.0. The file must first be saved as either an earlier version Paradox file or as a .dbf (dBase) file. However, the ODBC interface may be used to import data from Paradox 5.0 directly.

Why do I get strange values for my date variable when I import a SPSS portable file into STATISTICA?
    STATISTICA stores date values in the standard Julian format while SPSS uses a different format. To resolve the problem after importing the SPSS portable file into STATISTICA convert the values into the standard Julian format. Enter the following statement:
        =(v1-10010304000)/86400+1
    into the Long Name section of the Variable Specs dialog for the date variable.

I am unable to use the ODBC option with ms Access 7.0. What do I do?
    Windows 95 is shipped without an installation facility for ODBC drivers. You must obtain a special installation disk from ms. This file is also available from their BBS under the filename GE1263.exe. After you’ve extracted this file, you may use the following procedure to use your ms Access ODBC driver with STATISTICA functions.

1. First, make sure that you have installed the ODBC drivers that come with your STATISTICA software.
2. Reboot your computer.
3. Select the setup command associated with the ODBC program you extracted by executing GE1263.exe. You may do this through Windows Explorer (95) or File Manager (3.1).
4. Follow the installation procedure and finish it by selecting the Close button.
5. Again, reboot your computer.
6. You should now be ready to use 32-bit ODBC Access drivers from within STATISTICA.

I am trying to import a large ASCII file via Quick Import and the import fails. How can I import the file?
    Quick Import is limited to data files of small to moderate size. For large data files, you should use the Import Data—More option which is available only in the Data Management module.

Every time I try to import a certain Excel file into STATISTICA, I get a General Protection Fault error message ("This program has performed an illegal operation.") and STATISTICA crashes. What am I doing wrong?
    This behavior can also happen with other source applications. It is most likely that you are trying to import a corrupted or damaged file. First of all, you should try to import any other file and see whether the same behavior occurs. If this is the case, you might have to re-install the source application and/or STATISTICA.
    If the behavior is specific to one file, you should open the file in its source application (e.g. Excel), copy the contents into the Clipboard, paste it into an empty spreadsheet and save it under a new file name. Should the problem occur rather frequently, it's generally recommended to check your hard drive by using SCANDISK or advanced disk utilities.

Troubleshooting problems with Animated Overviews

When I run the Animated Overviews they seem to run fine from the CD, but I cannot run them from the Help menu. What is the problem?

We have received questions about this and the problem occurs for a few different reasons and applies to network server/client type installations.

During a server installation, STATISTICA copies the "multimed.ini" file to the network directory located on the server. In contrast, STATISTICA does not install the "multimed.ini" file to any of the clients.

To get around this when performing a Server/Workstation installation, we suggest to map a drive letter to the directory on the server that contains the STATISTICA installation files. Then, reinstall using the workstation setup directory, "wrkinst."

To map the drive letter, use these steps:

1. Open Windows Explorer
2. Click on Tools (located on the Windows Explorer title bar)
3. Next, select Map Network Drive.
4. Depending on where your server installation is located, assign a Drive letter and Path to your STATISTICA Installation directory

DATA MANAGEMENT

Cases, Variables, Formats

How to add/delete variables (columns of data):
    Variables may be added to and/or deleted from a data file either by (1) using the Drag-and-Drop operation ( insert to add columns), or (2) selecting an appropriate option from the spreadsheet menu of global operations on variables accessible by pressing the Vars button on the spreadsheet toolbar. For more information on those operations, see the section on Drag-and-Drop, and the spreadsheet toolbar.
    How to add variables before variable 1. The quickest way to add variables before variable 1 is to use the Insert action of the Drag-and-Drop facilities; you can also add variables after variable 1 and then move variable 1 after those variables.

How to add/delete cases (rows of data):
    Cases may be added to and/or deleted from a data file either by (1) using the Drag-and-Drop operation (insert to add cases), or (2) selecting an appropriate option from the spreadsheet menu of global operations on cases accessible by pressing the Cases button on the spreadsheet toolbar. For more information on those operations, see the section on Drag-and-Drop, and the spreadsheet toolbar.
    How to add cases before case 1. The quickest way to add cases before case 1 is to use the Insert action of the Drag-and-Drop facilities; you can also add cases after case 1 and then move case 1 after those cases.

How to select (highlight) the entire spreadsheet:
    Following the standard spreadsheet convention, click in the upper-left corner of the spreadsheet. This shortcut is useful, for example, when you intend to copy the entire file to the Clipboard. Note that by clicking twice in the upper-left corner, you will de-select (i.e., select and then de-select) the entire spreadsheet.

How to edit the contents of a cell (and not overwrite it):
    Double-click on the cell. To avoid clearing the contents of the cell (at the point when you start entering a correction), before you start typing, first double-click on the cell. This will enter the editing mode and will position the cursor within the cell.
    Press F2. Alternatively, you can follow the "old" spreadsheet convention and press the F2 key, which will also enter the edit mode for the currently highlighted cell.

How are dates represented in STATISTICA data files?
    Date values of variables are internally stored in Julian format, that is, as a single integer value that represents the number of days that have passed since January 1, 1900; for example a date entered and displayed as 1/21/1968 will be stored as the Julian date 24858; the (optional) decimals are interpreted as time (see the next topic). Date values stored in this manner can be used in subsequent analyses (e.g., in Survival Analysis in order to calculate survival times, see below) and transformed using arithmetic operations; at the same time, they can be displayed as dates in reports or graphs (e.g., used to label scale values).
    Julian date values can be displayed in the spreadsheet in numeric (Julian) format or in one of several pre-defined date display formats (e.g., 1/6/64, 6-Jan-64, Jan-1964, 01/06/64, 01/06/1964, 6-Jan).
    To change the date display format, select the Date format option in the Current Specs dialog (accessible by double-clicking on the variable name in the spreadsheet or from the spreadsheet Edit pull-down menu) and choose one of the pre-defined display formats.
    When entering dates into a new variable, you will first need to change the variable display format from Number (the default format) to Date in the Current Specs dialog and select the desired date display format. Now, you can enter the dates in any of the pre-defined date display formats (i.e., enter the dates in the format that is easiest to key in, even if it is different than the desired display format) and STATISTICA will recognize those formats, convert the display to the desired date format, and internally store the date values in Julian format. Please refer to the on-line Electronic Manual for more detailed information on entering new or pasting copied date values into the data file.
    You can create a single date variable from two (month, year or day, month) or three (day, month, year) variables as well as split a single date variable into two or three variables in the Date Operations dialog accessible from the Date Values option in the spreadsheet Vars pull-down menu.
    Please refer to the on-line Electronic Manual (click the Help button or press the F1 key in the Variable Specifications dialog) for examples of creating date variables from numeric variables and splitting date variables into numeric variables.

How is time represented in STATISTICA data files?
    Time values of variables are stored as (optional) decimal values representing the fraction of the day since midnight; for example, 6:00AM is stored as 0.25. Time values stored in this manner can be used in subsequent analyses and transformed using arithmetic operations; at the same time, they can be displayed as times in reports or graphs (e.g., used to label scale values).
    STATISTICA will format the display of time values according to the current settings in the International - Time Format option of the Windows Control Panel.
    Please refer to the on-line Electronic Manual (click the Help button or press the F1 key in the Variable Specifications dialog) for examples of creating time variables from numeric variables and splitting time variables into numeric variables.

How to use date-values as codes (e.g., as values of grouping variables):
    Some procedures in STATISTICA require that values of grouping variables (codes) be less than 32,000, i.e., less than the Julian values of some dates [this does not apply to grouping codes used in all Stats (and Quick Stats) Graphs procedures, Quick Basic Stats, Quality Control, Experimental Design, Process Analysis, and other procedures]. In order to use date values as codes smaller than 32,000, convert them into text values using the facility described below.

How to convert date-values into text values and vice-versa:
    In some circumstances it may be useful to create text values with date information (e.g., when using a date variable as a coding variable with codes smaller than 32,000, see the previous topic). In this case, you can transform the date variable into a variable containing date text values with numeric equivalents in a range that will allow them to be used as codes by all procedures of STATISTICA (i.e., numeric values less than 32,000). Use the Text Values/Dates dialog to convert dates into codes. This dialog is accessible by pressing the Text Values/Dates button in the Date Operations dialog.

How to review and edit variable specifications (names, formats, notes, formulas, etc.):
    Specifications of a single variable. Double-clicking on the variable name in the spreadsheet will open a variable specifications dialog where you can change the variable name, format, missing data value, long label, formula, or a DDE link for the current variable. (As in most other facilities commonly used in STATISTICA, alternatively, this dialog can also be accessed from the toolbar, flying menus called by the right-mouse-button, and the pull-down menus.)
    The dialog can also be used to access graphs, descriptive statistics, and a listing of sorted numeric and text values for the current variable (press the button Values/Stats or Graphs). From this dialog, you can also access the Text Values Manager (press the Text Values button) to review and change the assignments between the text and numeric values for the current variable.
    Specifications of all variables. You can also press the Variable Specifications button on the spreadsheet toolbar or the All Specs button in the dialog shown above), to bring up an editable, combined table of specifications of all variables in the current data file. The table format is convenient when you need to compare or edit specifications of several variables, especially when you need to paste and copy between variables (e.g., comments, formulas, or links), or extend a format definition or missing data code from one variable to subsequent variables (use the Fill Block Down option from the flying menu).

How are Missing Data handled in STATISTICA?
    Specifying Missing Data codes. A Missing Data code, that is, a value which signifies that there is no data for a particular case and variable (displayed as a blank cell in the spreadsheet) can be specified separately for each variable. To change the code, double-click on a variable name in the spreadsheet to access the Variable Specifications dialog for that variable. Alternatively, you can press the Combined Variable Specifications button on the spreadsheet toolbar to access a combined table for all variables (see above). The default Missing Data code in STATISTICA (used when new files are created, new variables are added, or data are imported) is -9999.
    Processing cases with Missing Data. The way in which missing data are handled when processing data can be adjusted individually for each analysis (see the field MD deletion in most analysis-definition dialogs). Whenever applicable, the user has the choice to eliminate them from calculations in a casewise or pairwise manner, substitute them with means, or reconstruct or interpolate them (e.g., in the Time Series module).
    Press the Help button or the F1 key in the respective analysis-definition dialog to learn about the specific missing data handling options available for the procedures of interest.

How to change the Missing Data code for individual variables:
    The value used to designate missing data values for individual variables can be changed in all variable specification dialogs (the default value is -9999; see also the section on reviewing and editing variable specifications).

How can I create a variable which contains the values 1 through n, in a random order without repetition?
    First, create a variable with random numbers between 0 and 1. You can do this with the spreadsheet formula =rnd(1). Then rank those numbers using the Rank option with sequential Ranks for Ties (this option will ensure that there is no repetition), available from the Vars toolbar button.

Can I use complex case selection conditions?
    Yes, Case Selection Conditions can be used to set user-specified conditions (of practically unlimited complexity), which define a subset of cases on which to base an analysis (without physically subsetting the data file).
    The Case Selection Conditions dialog can be accessed either by clicking on the Sel option from the Status Bar (on the bottom of the STATISTICA application window), clicking on the Select Cases button on the Startup Panel or an analysis definition dialog, or by choosing the Select option from the Options pull-down menu.
Logical conditions can be entered or edited in this dialog. Complex selection conditions can be created using logical operators and parentheses. For example,
        Exclude if: (v1<=10) OR (GENDER=’MALE’)
    This condition would exclude from an analysis those cases where the value of variable number 1 was less than or equal to 10 or those cases where the values of the variable GENDER equaled MALE.

How can I generate subgroup identifiers (codes) for my quality control charts?
    A spreadsheet formula can be used to generate subgroup identifiers. For example, the formula:
        =trunc((v0-1)/n)+1
where n is the number of times to repeat each value, can be used to generate the codes. For example, when n=4, then the following values (codes) of the grouping variable will be generated: 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3,...
    Note that the constant sample size option can also be used (instead of a grouping variable).

I have a data file (spreadsheet) with many (e.g., 400) variables. I would like to change the number of decimals that appear for all of the data. I know how to change the values for each variable, but is there a quick way to change the format for all variables?
    Yes, the number of decimals that appear for all of the data (or a range of variables) can be changed by pressing the toolbar Specs of All Variables (Table) button (or selecting the Vars toolbar button, and choosing All Specs), changing the format for the first variable of the range to be modified (e.g., 4.3 to 4.6 to add 3 decimal places), highlighting the desired range of variables in the Format column, clicking the right-mouse button, and choosing Fill Block Down. All formats in the selected block will be automatically modified.

How do I remove cases that have values greater than 2.5 (or some other multiple) of standard deviation for a variable (or set of variables)?
    You can write a STATISTICA BASIC program, using if then statements to find out which values are outside of +/-2.5 standard deviations and use the Delete command in Sequential Mode to delete those cases.

Text Values, Value Labels

How to enter/edit the assignments between numeric and text values:
    Normally, the assignments between text and numeric values are handled automatically as you enter or edit data in the spreadsheet (see the previous topic). However, in some circumstances, you may want to review, edit or restructure all assignments of values for a particular variable or copy text/numeric assignments from one variable to another. These operations can be performed in the Text Values Manager (accessible by pressing its respective button on the spreadsheet toolbar, by pressing the Text Values button in the variable specifications dialog for a particular variable; or from a flying menu (accessible by pressing the right-mouse-button anywhere on the respective variable in the spreadsheet).
    The Text Values Manager allows you to sort the assignments by text or numeric values, perform the Fill Down operation (to automatically reassign numeric values to text values), Copy numeric/text assignments from other variables or from case names, and perform other operations (press the Help button or the F1 key in this dialog to access the Electronic Manual).

How to switch between displaying text and numeric values in the data spreadsheet:
    As explained in the previous two topics, STATISTICA supports a double notation of values, where each value of a particular variable can simultaneously have a numeric and text identity. You can select a display of text or numeric values using a toggle switch button on the spreadsheet toolbar.

What are value labels?
    Value labels are comments or descriptions (up to 40 characters long) that can be assigned to specific text/numeric values in STATISTICA datasets. Each value of a variable can have a label assigned to it. Value labels can be displayed and edited directly in the spreadsheet (as well as in the Text Values Manager dialog). Value labels are automatically used in the output of all analyses which involve individual values (e.g., frequencies, crosstabulations, breakdowns, etc.).

How to display/edit the value labels:
    To display value labels in the spreadsheet, press the button while holding down the CTRL key or select the Text Values - Value Labels option in the spreadsheet View pull-down menu. Value labels can also be reviewed/edited in the Text Values Manager, accessible by pressing the Text Values Manager toolbar button.

How to copy a set of numeric/text value assignments and long value labels to other variables and other files:
    One of the options available in the Text Values Manager allows you to copy to the current variable the text values from other variables or from case names (see the Copy from box in the Text Values Manager dialog).
    When you need to copy the numeric/text value assignments and long value labels from one file to another, use the Merge files facility (in the Data Management module). One of the merge options available from the merge type selection dialog allows you to concatenate or replace text values with values from another file (see the topic on this merge option).

Clipboard and Drag-and-Drop operations in the spreadsheet
    When you copy or move a block in the spreadsheet, the values which are copied to the Clipboard will depend on the display mode of the spreadsheet. If the spreadsheet is displaying numeric values (or value labels) when the block is copied to the Clipboard, then only those numeric values (or value labels) will be copied to the Clipboard. If the spreadsheet is displaying text values when the block is copied to the Clipboard, then not only are the text values copied to the Clipboard, but also the corresponding numeric values and value labels (if any). This may result in the assignment of text values to numeric values that did not previously have text value equivalents.

I imported a file from Excel into STATISTICA. I want to change the default numerical assignment given to my Gender variable. How do I change "Male" from 100 to 1 and "Female" from 101 to 2?
    First, click on the Gender variable and then press the Vars toolbar button and select Recode. Under Category 1 type Gender=100 and set the resulting value to 1. Analogously, for Category 2, type Gender=101 and set the numeric value to 2. Then go back into the Text Values Manager (e.g. double-click on the variable name in the spreadsheet and select the Text Values button) and change the numeric values associated with "Male" and "Female" from 100 to 1 and 101 to 2, respectively. Click OK and the text values will be correctly associated with the new numeric assignments.

Operations on Blocks of Values

What are the Drag-and-Drop facilities?
    STATISTICA supports the complete set of standard spreadsheet (Excel-style) Drag-and-Drop facilities, as summarized below. For information on variable-speed scrolling while dragging a selection, and extending blocks outside the current window, please refer to the topic on variable speed highlighting.
    Moving a block. A selected block of data may be moved by pointing to the border of the selection (the cross-cursor will change to an arrow) and dragging it to the new location.
    Copying a block. A block may be copied by pointing to the border of the selection (the cross-cursor will change to an arrow) and pressing the CTRL key (a "plus" sign will appear next to the cursor) while dragging it to the new location.
    The default action which occurs when dragging a block without pressing the CTRL key may be changed from moving to copying in the STATISTICA Defaults:General dialog.
    Inserting a block. You may insert a block between rows or columns by pressing the SHIFT key while dragging the block and pointing the cursor between rows or columns (if you also press the CTRL key, the block will be copied and inserted instead of moved and inserted; a plus will appear next to the arrow-cursor as in the Drag-and-Drop copy operation, as shown in the illustration below).
    Clearing a block. A portion (or an entire) block can be cleared by dragging within a selected block (dimming the values to be deleted when you release the mouse).
    Extrapolating a block (AutoFill). Values within a block may be extrapolated by using intelligent, Excel-style, series extension facilities (see below).
    Numeric values are extrapolated using linear regression, text values are extrapolated following meaningful sequences (e.g., Q1, Q2, Q3, ..., or JAN, FEB, MAR, ...), see extrapolating blocks for more information.

How to rearrange blocks of data or ranges of cases and variables in a data file:
    There are three types of these operations: Clipboard-based, Global, and Drag-and-Drop-based; they operate differently and may produce different effects.
    Clipboard-based operations. Clipboard-based operations (invoked via the standard Clipboard keyboard, toolbar, or menu-commands) affect only the contents of blocks of data, rows, or columns, and they do not influence the overall size of the data file (e.g., they may empty a column but will not remove the column from the spreadsheet).
    The Global operations. The Global operations option (accessible from the pull-down menu or the spreadsheet Vars and Cases toolbar buttons) are performed on entire rows or columns as "units;" for example, they will move or delete entire columns and not only their contents.
    Drag-and-Drop operations. STATISTICA supports the standard (ms Excel-style) selection of Drag-and-Drop facilities, allowing you to easily perform both Clipboard-style (but without using the Clipboard) and Global (if the insert mode is invoked by pressing the SHIFT key) operations; for details see the previous topic.

How to expand a block in the spreadsheet outside the current screen:
    Blocks may be selected (1) by drag-selecting (holding down the left-mouse-button) with the mouse, (2) by clicking in one corner of the block to be highlighted, then scrolling to the desired opposite corner (the original cell will remain selected) and pressing the SHIFT key while clicking in that cell, or (3) by holding down the SHIFT key while using the cursor keys on the keyboard. To expand a previously-selected block, you may use the SHIFT-cursor key, or scroll the display with the mouse and press the SHIFT key while clicking in the desired corner of the block. To highlight a large block in "split-pane" mode (as shown below), click in a cell in one pane, then scroll to display the diagonally opposite corner in another pane and use SHIFT-click to select the block.
    Variable speed of block highlighting. Note that you can control the speed at which you scroll when you extend a block outside the current display window. By moving the cursor a short distance away from the spreadsheet (or Scrollsheet), you can scroll one line at a time when a block is selected; you can scroll one page at a time by moving the cursor further away from the spreadsheet (or Scrollsheet).

How to expand the data file size by using Drag-and-Drop:
    If you drag a block past the current boundary of the spreadsheet, the data file will expand to make room for the new data. You can also expand the spreadsheet by using the Insert operation. This operation will insert the block between cases or variables, thereby increasing the size of the spreadsheet. Note that whether you move, copy, or insert new cases or variables, when the block contains only part of a variable or case, then STATISTICA will fill the remaining values in the variable/case with missing data.
    Note that if you work with large data files (e.g., more than 10 megabytes), you may wish to enable the option to Ask before inserting rows or columns in the STATISTICA Defaults: General dialog.

Splitting and Merging Files

Can I merge the long value labels (or text/numeric value assignments) from two files?
    Yes, when you select the Merge files option in the Data Management module (see above), in addition to the Merge Cases and Merge Variables, you will be able to choose the Merge Text Values option. In the subsequent dialog you will be able to select the way in which the text values from the two files are to be merged.

Formulas, Data Transformations, Recoding, Verifying, Cleaning Data

How to calculate (transform) values of a variable:
    Spreadsheet formulas. To perform single-line data transformation and recoding operations on single variables, you can use the data spreadsheet formulas. Double-clicking in the spreadsheet on the name of the variable which you want to transform will open the Variable Specification dialog (see below) where a data transformation or recoding formula can be entered directly into the Long name (label, link, or formula) field.
    Following the Windows spreadsheet formula conventions (e.g., ms Excel), start the formula with an "=" (otherwise STATISTICA will not recognize that the text is to be interpreted as a formula). For example, enter =(v1+v2)/2. Variables can be referenced by their names (e.g., Income, profit, TEST1) or numbers (e.g., v1, v2, v3, ...); v0 is the case number. Logical operators can be used to define conditional transformation expressions.
    Press the Functions button when you are in this dialog to access the Functions Wizard facility that can be used to select functions and other elements of the syntax (for the formulas) and insert them into the formula editor. The Wizard also allows you to quickly access information on any of the functions and the complete on-line syntax documentation.
    After entering a formula and pressing OK, you are given the option to recalculate the variable now. If you choose not to recalculate at this time, you can do so later by pressing the spreadsheet toolbar Recalculate button or the Recalculate All key (F9).
    The STATISTICA BASIC programming language. If you need to write more complex data transformation programs than those which can be entered via spreadsheet formulas, an integrated programming language (STATISTICA BASIC) can be used. The language can be accessed from the pull-down menu Analysis in each module.
    STATISTICA BASIC is a comprehensive programming language, supporting loops, nested conditional statements, compound instructions, multidimensional arrays, matrix operations, random access to individual records, custom-designed recoding functions, calls to external DLL procedures, allows you to access STATISTICA graphics facilities, etc.
    The language comes with an integrated environment that allows you to write, edit, debug, and execute your programs (you may open multiple STATISTICA BASIC programs in separate editor windows in the same module and copy and paste between them). The central part of the BASIC environment is a flexible Function Wizard facility that allows you to select the respective functions or keywords and insert them into the program.
    The syntax of STATISTICA BASIC is very simple. While writing your programs, you can access Examples (shown below) and a concise Syntax summary by pressing the respective toolbar buttons (see the Help:Examples and Help:Syntax buttons). It is convenient to have the help screen available while you write your programs; you can keep the Help window always on top of the STATISTICA window if you press the Always On Top button in the Help window.
    For more information on STATISTICA BASIC, please refer to the STATISTICA BASIC section.

Can STATISTICA automatically recalculate all spreadsheet formulas when the data change?
    Yes, by setting the option Auto-recalculate when the data change in the Recalculate dialog (accessed by pressing the Recalculate button in the toolbar), all spreadsheet formulas will be automatically recalculated when the data are changed in the spreadsheet. Alternatively, you can press F9 in the program in order to manually recalculate the formulas in the spreadsheet.

How to generate values from a normally distributed random variable?
    There are several ways in which to do this. You may use the spreadsheet formulas to fill a variable with random values from a normal distribution with a given mean and standard deviation. This is accomplished by using the inverse function for the normal distribution (vnormal(x)). As an example, to fill a variable with values from a normal distribution with a mean of 50 and a standard deviation of 3, type the following formula in the long name edit field of the current specs dialog for the variable:
    =vnormal(rnd(1),50,3)
The inverse function can also be used in a STATISTICA BASIC program. Additionally, you can generate values fom the standard normal distribution (mean = 0, standard deviation = 1) by using the normal(x) function in STATISTICA BASIC.

What is the simplest way to recode values of a variable (e.g., split a continuous variable into categories)?
    Recoding functions of practically unlimited complexity can be custom-defined in STATISTICA BASIC and used repeatedly in your data transformation programs. However, a quick on-line recoding facility can also be accessed directly from the spreadsheet at any point (see Recode option, accessible from the spreadsheet Vars toolbar button). The scrollable Recode Values dialog which will be displayed allows you to define new values of the current variable depending on the specific conditions, which you define.
    When specifying the conditions, follow the standard syntax conventions which are common in STATISTICA to all those procedures which involve any operation of "selecting cases" based on their values. For example, the recoding conditions specified above would "translate" the negative values of the current variable (Measur1) into -1 and positive values into 1. Note that the 0's (the only value which is not included in the two recoding conditions) would be left unchanged, as set using the radio button unchanged (see the box Other in the lower right corner of the dialog). You can also use here Case Selection Conditions that have been specified somewhere else and saved to disk files (either as individual conditions or sets). Note that recoding conditions may be much more complex (see the Electronic Manual by pressing the Help button or the F1 key in this dialog) and they can be defined such that the new values of the current variable do not depend on the old values of that variable, but only on values of some other variables in the dataset. Thus, this facility can be used not only to recode existing data, but also to create values of a new variable based on conditions met by other variables (as illustrated in the next topic).

How to create values of a new variable based on conditions met by other variables:
    You can use any of the data transformation facilities: spreadsheet formulas or STATISTICA BASIC. However, often the quickest way to do it would be to use the on-line data recoding facility described in the previous topic, which is accessible at any point from the data spreadsheet (see the spreadsheet Vars toolbar button). As mentioned in the previous topic, the currently highlighted variable does not even have to be included in the text of the recoding conditions. Thus, you can use this facility to create values of a variable based on conditions met by other variables.
    For example, you can add a new (empty) variable to the data file, and then use this facility to create the new values. For instance, the recoding conditions could be used to assign 1's to the new variable for all "male subjects, 18 to 25 years old with cholesterol levels below 200;" 2's -- to "male subjects 18 to 25 years old with cholesterol levels above 200;" and assign the missing data value to all other subjects.

How to verify and "clean" data:
    An interactive data-verification and cleaning facility is provided in the Data Management module. In order to verify data, bring up the Data Management module, and select the option Verify Data Values from the startup panel or the pull-down menu Analysis. The Verify Data dialog which will appear allows you to enter the conditions to be met by the data.
    Follow the standard syntax conventions which are common in STATISTICA to all those procedures which involve any operation of "selecting cases" based on their values. You can also use here Case Selection Conditions that have been previously saved to disk files (either as individual conditions or sets). Pressing the Help button or the F1 key in this dialog will bring up a comprehensive description of all verification options with examples.
    The verification can be as simple as checking whether values in a variable are "legal" (e.g., only 1 and 2 might be allowed for Gender) or whether they fall within allowed ranges of values (e.g., Age must be more than 0 and less than 200). It can also be as complex as checking multiple logical conditions that some values must meet in relation to other values.
    Consider the following example of conditional verification:
    If a person is a male or less than 14 years old, then the number of pregnancies for that person cannot be more than zero.
    In order to apply these conditions, you would specify (for example):
    Invalid if: (v1='MALE' or AGE<14) and PREGN>0
    After you specify the verification conditions, press the OK (Verify) button and the dataset (or the selected range) will be tested sequentially (one case at a time) for its consistency with the set of conditions which you have specified.
    When a case is found which does not meet the conditions, then the respective row of data in the spreadsheet will be brought up and highlighted and the Data Verification dialog will open allowing you to either ignore the inconsistency (and continue or stop the verification) or edit (correct) the case.
    Now you can press Review/Edit the Data (or press Enter) to edit the case which is highlighted in the spreadsheet. Alternatively, you can Ignore and Continue or Ignore and Exit the verification. Note that if you decide to edit the case, then after you are done, pressing the floating Cont (continue) Verify Data button in the lower left corner of the screen will continue the verification.

Custom data verification and cleaning facilities:
    For some (probably less common) applications, you can easily develop interactive data cleaning and verification applications using STATISTICA BASIC. A single statement (PauseAndEdit) can be used to bring up an interactive dialog allowing you to review and edit any variables in the data file.

How to perform a multiple sort:
    Use the Sort option from the startup panel (or the pull-down menu Analysis) of the Data Management module. This will bring up a dialog where you can specify the key variables and the type of sort. If you need to sort based on more than 3 keys, press the More Keys... button to switch to a larger dialog.

How to rank-order values of a variable (replace values with their ranks):
    Select the Rank option by pressing the Vars button on the spreadsheet toolbar to bring up the Rank Order Values dialog, then press the Variables button to select the variables to be ranked. Optionally, you can also specify a subset of cases to be affected by the operation, use case weights, and set a number of options to perform particular (non-default) types of ranking (press the Help button or the F1 key in this dialog to access the Electronic Manual).

How to transpose data (convert cases into variables):
    Use the Transpose option accessible from the data spreadsheet Edit menu. A hierarchical menu will allow you to select either the Block or Data file transposing option.
    Transposing a block of data. The block transposing operation will affect only the contents of cells in the currently selected block highlighted in the data spreadsheet (the block must be square); the variable names and case names will not be affected.
    Transposing a data file. The data file transposing option will restructure the entire file. Note that when you transpose a file, case names become variable names and variable names become case names (for details, press the F1 key to access the Electronic Manual).

How to automatically fill ranges of data in the spreadsheet:
    In addition to the custom-defined operations of filling specific ranges of data with values, available via the data transformation options, STATISTICA can fill selected blocks with random values, and also supports standard (ms Excel-style) block extrapolation operations, which can be invoked by dragging a border of the block onto the adjacent area of the spreadsheet that is to be filled following the extrapolation rules.
    Random value fill. Select the Fill Random Values suboption of the Fill/Standardize Block option in the spreadsheet flying menu (click the right-mouse-button). The currently highlighted block will be filled with random values (following a uniform distribution) in the range of 0 to 1.
    Extrapolation of values (AutoFill). A selected series of values (consisting of at least two values) will be extended using linear regression to extrapolate values (see also the next topic). If a single value (or a repeated value) is selected, the value will be copied into the extrapolated block; also sequences consisting of names of months, days, or quarters can be automatically extended.
    Fill Down or Right. Finally, simple spreadsheet-style "fill block" (Down or Right) options are available which fill the currently selected block by duplicating the first row or column (respectively) of the block. To use this facility, first, select a block to be filled in the spreadsheet. Then, use the Fill/Standardize Block option accessible from the data spreadsheet Edit menu and the spreadsheet flying menu. A hierarchical menu will allow you to select either the Fill Down or Fill Right option.
    The former will copy (duplicate) the first row of the block to all remaining rows of the block; the latter will copy the first column of the block to the remaining columns of the block. Both options work in a manner similar to ms Excel's Fill Right and Fill Down facility (also available in Excel from the Edit menu).

How does the extrapolation of blocks (AutoFill) work?
    A horizontal or vertical series in a block can be extrapolated by dragging the block Fill Handle (a small, solid square located on the lower-right corner of the block). This is illustrated in the example in the previous topic. STATISTICA can create series of values such as sequential numbers, linear extrapolations and dates (e.g., you can extend a series such as 1, 2, 3 to include 4, 5, 6, ...).
    You can extrapolate a block in the following ways:

• If the initial selection contains repeated values, these values will be duplicated in the extended block.
• Non-identical values are extended by linear regression to compute extrapolated values for the series.
• If the block contains names of days (e.g., MON, TUE, WED, ...), months (e.g., JAN, FEB, ...), or quarters (e.g., Q1, Q2, ...), then STATISTICA will extrapolate the rest of the names in the series (e.g., MAR, APR, MAY, ..., DEC; THR, FRI, SAT, SUN; Q3, Q4, respectively).
• If the block contains other text values, then the text value of the last cell will be copied into the extended block.

How to standardize values in a block:
    Highlight a block of values in the spreadsheet, then from the spreadsheet flying menu (click the right-mouse-button) select either Standardize Rows or Standardize Columns. The standardized values are computed as follows:
    Std. Value = (raw value - mean of highlighted row/column) / std. deviation
    You can also standardize selected variables via the option Standardize Variables in the Data Management module (that procedure works independent of the currently selected block but takes into account the current case selection conditions and weights).

I have noticed that when I use an ArcSin transformation in a STATISTICA spreadsheet formula, the results are given in radians, not degrees. Is there anyway I can change this?
    While there is no option which allows toggling between the two types of results, multiplying the spreadsheet formula by 180/Pi will convert the radians into degrees.

Analyses, On-line Statistics

How to select variables for an analysis:
    Every analysis definition dialog in STATISTICA contains at least one Variables button which allows you to specify variables to be analyzed. You can click on it (or press V). If you forget to specify variables and press OK to start the analysis, STATISTICA will ask you for the variables to be analyzed. The Variable Selection window which will appear supports various ways of selecting variables and it offers various shortcuts and options to review the contents of the data file.

Can variables be selected for analyses by highlighting them in the spreadsheet?
    Yes, this shortcut method is supported. If you select a block in the data spreadsheet, then the variables included in the block will automatically become pre-selected for the next analysis. Note that this shortcut is designed to limit the chance of producing unintended results in the following ways:
    The pre-selection of variables by marking a block in the spreadsheet works only as long as you have not selected a specific list of variables for the analysis (i.e., it will never overwrite your previous choices; STATISTICA will remember your previous choices).
    Also, if the variables from the block are not what you intend to analyze, you do not need to "undo" the selection: When you enter the Variable Selection window, the list of pre-selected variables will be highlighted, thus, the first click of the mouse will de-select the previous range (unless you keep the CTRL key pressed). If you prefer to use the keyboard to specify the list, then the first (non-cursor moving) key you press will delete the previous entry in the variable selection edit field.

How to select a subset of cases (observations) to be included in an analysis:
    Before an analysis begins (i.e., before the data are processed), you can instruct the program to select only cases (i.e., rows in the spreadsheet) which meet some specific selection criteria. A facility to define and manage such Case Selection Conditions can be accessed from the pull-down menu Options, by pressing F8, or simply by double-clicking on the status bar field Sel: OFF which normally shows the current status of the Case Selection Conditions (when disabled, the field reads Sel: OFF, when enabled, the field reads Sel: ON). For a complete reference, access the Electronic Manual when the Case Selection Conditions window is displayed.

How are the Case Selection Conditions stored/saved?
    Currently-specified Case Selection Conditions (definitions of subsets of data) are automatically stored by STATISTICA when you change modules. You can also save and open them (and maintain libraries of Case Selection Conditions) by using the Open or Save As buttons in the Case Selection Conditions dialog. The list of supplementary files associated with the current dataset (maintained in the Workbook dialog) offers easy access to those libraries, and the list is accessible from every Open/Save dialog.
    Note that the saved Case Selection Conditions can also be used in all STATISTICA facilities which allow you to define subsets of cases. For example, the same conventions apply to the data Recoding facility (available by pressing the spreadsheet Vars toolbar button), data verification facilities (available in the Data Management module), or all procedures which allow you to custom-define multiple subsets of data (such as frequency tables, or multiple-subset scatterplots).

What is the quickest way to review basic descriptive statistics for a variable?
    You can either use the Quick Basic Stats option accessible from the flying menu in all spreadsheets and Scrollsheets, or the data listing option Values/Stats in the Quick Stats Graphs menu in the following manner. Highlight any cell in the desired variable in the spreadsheet (or Scrollsheet), then press the toolbar button Quick Stats Graphs to bring up the menu of statistical graphs and options. Alternatively, press the right-mouse-button and select Quick Stats Graphs from the flying menu. Then double-click on the Values/Stats of... option (or highlight this option and press Enter). The resulting window will display information about the selected variable, a sorted list of its values, and descriptive statistics. The descriptive statistics can be copied to the Clipboard by pressing the Copy button (see the lower part of the window).
    Note that a more extensive selection of descriptive statistics and other basic statistics (correlations, frequency tables, distributions, analyses by..., etc.) can be requested from the Quick Basic Stats menu available from the toolbar button or right-mouse-button flying menus in all spreadsheets and Scrollsheets and from all pull-down Analysis menus.
    For information about a facility to produce descriptive statistics for columns or rows of the currently selected block of values, see Block Stats.

What is the quickest way to review ordinal descriptive statistics (median, quartiles) for a variable in the spreadsheet or Scrollsheet?
    You can either use the Quick Basic Stats option accessible from all toolbars, or the flying menu in all spreadsheets and Scrollsheets, or the option Box-Whisker in the Quick Stats Graphs menu. The latter method will also visualize the statistics in a graph. Highlight any cell in the desired variable in the spreadsheet (or Scrollsheet), then press the toolbar button Quick Stats Graphs. Alternatively, press the right-mouse-button and select Quick Stats Graphs from the flying menu which will pop up. Either method will bring up the menu of statistical graphs and options.
    Depending on what kind of descriptive statistics you would like to review, select one of the Box-Whisker of... options for a single variable (e.g., to review ordinal descriptive statistics and the range, select the option Median/Quart/Range): The graph will appear on the screen, including the specific values of the respective descriptive statistics.
    For a more complete description of the distribution of the variable, you can paste into the graph the basic descriptive statistics (copied from the dialog described in the previous topic). Note that many other statistical graphs of distributions of variables (e.g., a large selection of user-defined, univariate and multivariate box and whiskers plots with outliers) can be requested from the Stats Graphs menu (available in the pull-down menu Graphs).

What is the quickest way to review a sorted list of all values of a variable?
    Make sure that the cursor is (anywhere) on the desired variable in the spreadsheet (or Scrollsheet). Press the Quick Stats Graphs button on the toolbar. From the Quick Stats Graphs menu (see the previous two topics), select the Values/Stats of... option. A sorted list of all numeric- and corresponding text-values (if there are any) for that variable will appear.
    Note that the assignments between the numeric and text values for the variable can be edited in the Text Values Manager. To produce a frequency table of values, use the Quick Basic Stats facility (see the next topic).

What are Quick Basic Stats?
    The Quick Basic Stats options are available from either the spreadsheet or Scrollsheet toolbar button, the right-mouse-button flying menu, or from the Analysis pull-down menu of any module of STATISTICA. Quick Basic Stats can be invoked at any point of your data analysis (e.g., to provide supplementary information when you review output from any STATISTICA module).
    These procedures include a selection of basic statistics which can be performed on long lists of variables (e.g., correlation matrices for all variables in the dataset), and all analyses can be performed by groups, for every value (i.e., code) of a selected grouping variable.
    One of the advantages of this facility is that usually you do not have to select variables for Quick Basic Stats analyses via dialogs; instead, the variables are automatically selected from the currently highlighted block in the spreadsheet or Scrollsheet.
    For more information, press the Help button or the F1 key to access the Electronic Manual; see also the next topic.

What are Block Stats?
    Statistics for each row or column in a selected block may be computed and added to the spreadsheet or Scrollsheet (i.e., appended at the end) by selecting the desired Block Stats from the flying menu (or from the Edit pull-down menu). The selected statistics will be added to the end of the data file or Scrollsheet as additional rows or columns. The variable name will contain the name of the selected statistic and the long name will include the range description.
    If you select Block Stats/Columns, the results of the selected statistic will be appended as a new case and the case name will contain the name of the selected statistic and the range description.
    In addition, Block Stats Graphs are also available from this flying menu to produce statistical graphs for the values in each row (if Block Stats/Rows was selected) or each column (if Block Stats/Columns was selected) of the currently highlighted block in the spreadsheet.

Can matrix data (e.g., correlations or covariances) instead of raw data be used for input?
    Yes, STATISTICA supports a variety of matrix-file types as input data (e.g., for Regression, Factor Analysis, Reliability, and other modules). Matrix files can be edited in the spreadsheet (as if they were raw data files), however, in order to be properly interpreted as matrices by the program, they need to meet specific content and format conditions (depending on the matrix type). STATISTICA also supports the multiple matrix data file format for input in procedures (such as Structural Equation Modeling) which support matrix data for groups.
    Refer to the respective module of STATISTICA for details on matrix file input (press the Help button, the F1 key, or use the Electronic Manual for specific descriptions of the matrix file formats supported by STATISTICA).

How do I create a control chart for a data file where each column represents an observation in a sample? For example, I have 20 samples, each of size 5, so my data file has 20 rows and 5 columns.
    You can use the Block Stats/Rows option by right mouse clicking on the desired row to compute columns of means and standard deviations. Add a column using the Vars toolbar button and fill in the sample size with a spreadsheet formula (e.g. =5). Then use the means, standard deviations, and sample sizes as the input instead of the raw data.

Import, Export, DDE

How to access data from Excel and other foreign data files:
    Clipboard. The quickest, and in many cases easiest way to access data files from other Windows applications (e.g., spreadsheets) is to use the Clipboard, which in STATISTICA supports special Clipboard data formats generated by applications such as ms Excel or Lotus for Windows. For example, STATISTICA will properly interpret formatted cells (such as 1,000,000 or $10) and text values.
    File import facilities. Data files from a wide variety of Windows and non-Windows applications can also be accessed and translated into the STATISTICA format using the file import facilities which also include access to virtually all databases (via STATISTICA's flexible ODBC support) and options to access formatted and free format text (ASCII) files.
    The main advantages of using the file import facilities (over the Clipboard) are that:

• They allow the user to specify the exact way in which the translation is to be performed (e.g., access named ranges in the foreign data files, decide whether or not to import variable names, text values, and case names, and how to interpret them); and
• They allow the user to access types of data which are not (or not easily) accessible to Clipboard operations (such as long value labels and special missing data codes).

How to export data from STATISTICA to Excel and other foreign data files:
    The Clipboard and data file translation facilities described in the previous topic in the context of importing foreign file formats can also be used to export data from STATISTICA to other formats. The same selection of formats and data types is supported when importing data to and exporting them from STATISTICA (see the previous topic).

Can STATISTICA access data from mainframe databases?
    Yes, a flexible ODBC interface (including options to combine fields from multiple tables) is provided via the Import Data facilities, available in the File pull-down menu in every module.

Does STATISTICA support ODBC?
    Yes, via the Import Data facilities, available in the File pull-down menu in every module. The STATISTICA ODBC interface includes options to combine fields from multiple tables and provides flexible access to a wide variety of database management files, including both mainframe and microcomputer formats (e.g., dBASE for Windows, Paradox, Sybase, Oracle, SAS, etc.).

How to set up DDE (Dynamic Data Exchange) links between STATISTICA and files created by other Windows applications:
    You can establish DDE (Dynamic Data Exchange) links between a "source" (or server) file (e.g., a ms Excel spreadsheet) and a STATISTICA data file (the "client" file), so that when changes are made to the data in the source file, the data will be automatically updated in the respective part of the STATISTICA spreadsheet (client file).
    A common application for dynamically linking two files would be in industrial settings, where the STATISTICA data file would be dynamically linked with a measurement device connected to the serial port (e.g., in order to automatically update specific measurements hourly).
    DDE links can be established using the quick, "paste-like" Paste Link option in the spreadsheet pull-down menu Edit, or by entering a definition of the link into the Long name (label, formula, link): field of the Variable Specifications dialog. When a link is established, it can be managed using the Links Manager (accessible via the Links... option in the spreadsheet pull-down menu Edit).
    Refer to the Electronic Manual (accessible by pressing the Help button or the F1 key from either of the two options) for details and examples. There are also other applications for the DDE data integration facilities in STATISTICA; see the Electronic Manual on DDE for details.

DDE with Excel 7.0: Troubleshooting
    The statement defining the link for Excel 5.0 is specified as application, file, and range, and as part of the range you can specify in which sheet the data are located. However, the Excel 7.0, DDE server may produce a nonstandard statement (it includes a different specification of the sheet, and the link cannot be established; i.e., the Paste Link option in the Edit menu does not produce a working DDE link).
    The first fix is to edit the statement (generated by Excel 7.0) that defines the link (in STATISTICA) by deleting out the nonstandard specification of the sheet. For example, if the DDE server creates the following link,
@Excel|C:\EXCEL\[DATA.XLS]Sheet1!R1C1:R11C1
you will need to delete the reference to Sheet1 in order for the link to be established. The link would then be
@Excel|C:\EXCEL\DATA.XLS!R1C1:R11C1
    You can edit links in the Current Specs dialog (accessible by double-clicking on the name of the variable, to which the link is to be established), the All Specs dialog (accessible via the All Specs spreadsheet toolbar button) or the Link Manager dialog (available by selecting Edit..Links from the Edit pull down menu.
    The limitation here is that you can only create a link with the first sheet in any Excel 7.0 file. If your data are in sheet one, this is not a problem. If the data are not in sheet one, you would need to link the data to a new Excel 7.0 file and place them in sheet number one.

Why do I get an error message when trying to import a Paradox (.db) file?
    The STATISTICA Quick Import facility does not support files from Paradox, release 5.0. The file must first be saved as either an earlier version Paradox file or as a .dbf (dBase) file. However, the ODBC interface may be used to import data from Paradox 5.0 directly.

Why do I get strange values for my date variable when I import a SPSS portable file into STATISTICA?
    STATISTICA stores date values in the standard Julian format while SPSS uses a different format. To resolve the problem after importing the SPSS portable file into STATISTICA convert the values into the standard Julian format. Enter the following: =(v1-10010304000)/86400+1 into the Long Name section of the Variable Specs dialog for the date variable.

I am unable to use the ODBC option with ms Access 7.0. What do I do?
    Windows95 is shipped without an installation facility for ODBC drivers. You must obtain a special installation disk from ms. This file is also available from their BBS under the filename GE1263.exe. After you’ve extracted this file, you may use the following procedure to use your ms Access ODBC driver with STATISTICA functions.

1. First, make sure that you have installed the ODBC drivers that come with your STATISTICA software.
2. Reboot your computer.
3. Select the setup command associated with the ODBC program you extracted by executing GE1263.exe. You may do this through Windows Explorer (95) or File Manager (3.1).
4. Follow the installation procedure and finish it by selecting the Close button.
5. Again, reboot your computer.
6. You should now be ready to use 32-bit ODBC Access drivers from within STATISTICA.

I am trying to import a large ASCII file via Quick Import and the import fails.
    Quick Import is limited to data files of small to moderate size. For large data files, you should use the Import Data—More option which is available only in the Data Management module.

I am trying to establish a DDE link with another dataset and I get a "DDE Init Failed" error message. The link is not established.
    One of the steps in the DDE procedure is to ask the source to give up the data.
    STATISTICA cannot wait for ever for a reply in case there is a problem because this would hang the system so there is a time limit. The default for this is 1 second.
    On a fast computer this is not a problem, but on slower machines such as a 75MHz, sometimes (especially when STATISTICA is still loading in the background on start-up) the computer has not linked in that first second.
    The solution is to go into STATIST.INI and add the following...
        in the [Statistica] section:
        DDEDelay=5000
    where 5000 = 5 seconds because it is in milliseconds (use another number as necessary).

General

What are Workbooks?
    Workbooks help to organize sets of supplementary files (e.g., Scrollsheets, graphs, text/graphics reports, user programs, recodes, etc.) which have been created or used (e.g., reviewed) during the analysis of a dataset. The STATISTICA Workbook facility maintains a list of all files used with the current dataset.
    An updated list of these files is automatically saved with the data file. Selected files in the list can be marked Auto Open, they will then be automatically brought up whenever you open the current dataset.

Can I open more than one input data file simultaneously?
    Different data files can be used simultaneously for input in different modules (as mentioned before, different modules can process the same or different data files). From the File Manager, data files may be opened directly (into the STATISTICA File Server application, see below) by double-clicking on them. Each data file will be opened into a separate File Server application window (see the next topic); in this manner, you may simultaneously open as many data files as you wish.
    In one module, only one file (i.e., Workbook) at a time can be used as the main input data file and reside in the data spreadsheet; other files can be opened simultaneously in one module for reference only, thus, they can be opened only as Scrollsheets (if they were previously saved into Scrollsheet files). Refer to the section on differences between spreadsheets and Scrollsheets for more information.
    Multiple Megafile Manager files can be opened simultaneously in the Data Management module, allowing you to easily move data from standard STATISTICA data files to different aggregated (or archival) databases, or to create STATISTICA data files from data stored in various Megafile Manger databases.

What is the STATISTICA File Server?
    This small, rapidly-loading module of STATISTICA provides facilities for accessing and modifying all STATISTICA files, including data files (file name extension *.sta), Scrollsheets (*.scr), graphics (*.stg), and STATISTICA BASIC files (*.stb). STATISTICA files may be opened from the Windows File Manager by simply double-clicking on the file name. One of the advantages of using the STATISTICA File Server application is that you may have as many instances of the File Server open at once as your operating system will allow, permitting you to review many data files at one time (see the previous topic).
    In addition, the STATISTICA File Server application provides OLE support for STATISTICA files in other applications (if there is not any other module of STATISTICA currently running).
    Custom-designed, new modules of STATISTICA. The File Server can also be used as a foundation to develop new STATISTICA modules (using STATISTICA BASIC) with their own startup panels, and Scrollsheet and graph output.

Does STATISTICA support Drag-and-Drop opening of data files and graphs?
    Yes. In the Windows File Manager you may simply drag any STATISTICA data file (file name extension *.sta), Scrollsheet file (file name extension *.scr), or graphics file (file name extension *.stg) onto the icon for a STATISTICA module in order to automatically open the file in the respective STATISTICA module.

How much disk space is necessary to perform database management operations?
    In order to allow the user to revert back to the original data file after file editing, STATISTICA creates temporary and backup files. Thus, in order to edit a data file, the program will need at least twice as much free disk space as the size of the file to be edited. More disk space is necessary if you perform large-scale editing operations (e.g., changing large blocks) on very large data files: to allow for the Undo function, STATISTICA will then need to store multiple copies (up to 16) of the modified sections of data. Some operations (e.g., import/export) use intermediate scratch files, thus they may need additional space.

In what circumstances does STATISTICA issue the "Disk full" message?
    STATISTICA uses the disk drive which is designated for temporary files in your Windows configuration for its temporary files. On some systems, this disk may be a RAM drive of a relatively small size. The limited space on that drive may be quickly used up (by STATISTICA and other Windows applications) and the Disk full message issued even though there is still free space on some other drive. To check the current setting, look for the SET TEMP= line in your autoexec.bat file (in the root directory of the boot disk). To remedy the situation, set the location for temporary files to your largest disk.
    Note also that because the ms Windows system and various Windows applications use large temporary disk files, the actual amount of disk space effectively available to an application may be much smaller than the amount reported by File Manager immediately after starting the computer.

What are the differences between the data management options accessible in every module and the Data Management module?
    Most of the commonly-used data management facilities are integrated with the spreadsheet and available from the data spreadsheet in every module (either from the toolbar, pull-down menu Edit, or the flying menus). These commonly-available facilities include all spreadsheet operations on cases and variables, transposing data, data transformations through the spreadsheet formulas, the STATISTICA BASIC programming language, and the on-line recoding facilities, ranking, filling ranges, shifting, DDE, management of text values, date values, long value labels, and many other options (including direct import/export functions and import via ODBC).
    The Data Management module includes all of these options and additional, specialized data management facilities such as relational merge, interactive data verification and cleaning, specialized import/export facilities (e.g., allowing import/export of double notation of values to file formats which normally do not support double notation), and others.
    Access to Megafile Manager facilities. In addition, the Data Management module of STATISTICA provides access to all facilities of the Megafile Manager database system (see the next topic), which offers options to access and process data in unusual data formats (e.g., data organized into extremely large records or data with very long text values). A unique feature of Megafile Manager is that it can process data with extremely long records (up to 8 megabytes per record). Megafile Manager also can be used as an archival data base system to store data combined from various sources (preserving their original formats). Easy to use (one step) and flexible facilities are provided in the Data Management module to move data in and out, between STATISTICA data files and archival Megafile Manager databases.

What is Megafile Manager?
    Overview. Megafile Manager is a specialized data base management system accessible from the Data Management module of STATISTICA (see the previous topic). Its unique feature is that it can manage and directly process types of data which need to be transformed, aggregated, extracted, or cleaned before they can be directly accessed by any STATISTICAl or graphics procedures of STATISTICA (e.g., data organized into very long records or data embedded inside very long text values).
    Megafile Manager can process extremely large records of data (rows): up to 32,000 columns with up to 255 characters each (up to 8 megabytes per row). Data organized in such long records can, for example, be produced by some automated quality control measurement devices or other data acquisition or monitoring equipment. Also, such files are sometimes useful in maintaining integrated, large archival data banks consisting of numerous merged or concatenated files.
    Megafile Manager and other STATISTICA facilities. Most likely, the majority of common data processing needs can be easily addressed with the standard procedures available in every module of STATISTICA. Therefore, Megafile Manager will typically be used only when there is a need for facilities to handle very unusual data importing, management, and pre-processing needs, for example, in order to maintain very large databases, or perform pre-processing of imported long text values, etc. Note that the standard procedures (offered in every module of STATISTICA) can also handle large size and very complex tasks; for example:

• to compute descriptive statistics or correlation matrices for very many variables (e.g., over 1,000) in a single run, use the Quick Basic Stats facilities available in every module via the Analysis pull-down menu or the toolbar Quick Basic Stats button;
• to perform complex recoding and aggregation of data, extensive operations on text values, or complex cleaning and verification of data, use STATISTICA BASIC, available in every module of STATISTICA from the Analysis pull-down menu;
• to import large files from other applications (e.g., Excel, Lotus, Quattro Pro, Paradox, etc.), use the Import options available from the pull-down menu File in every module of STATISTICA;
• to perform sequential, hierarchical, or relational merging of files, use the Merge option from the Analysis pull-down menu in the Data Management module, or use the functions available in STATISTICA BASIC for accessing external files (and then use the available powerful STATISTICA BASIC functions to merge data via simple or complex user-defined algorithms).

Can I Undo editing operations?
    When you click the Undo toolbar button or select Undo from the spreadsheet Edit pull-down menu, you will be able to undo most operations such as editing, moving or copying blocks, random fill, recoding or ranking variables, etc. Multi-level undo is supported (with 16 buffers), so you can undo up to 16 of the most recent spreadsheet operations.

How to restore an inadvertently overwritten data file:
    When you are overwriting an existing data file (while performing the Save As... operation), STATISTICA will give you an option to create a backup (under the same file name but with the file name extension *.bak). As long as you do not use Save As... twice in a row (which will overwrite the backup), you can return to the file as it was before being saved by using the backup.

What is the difference between the Output header and the File headers?
    Output header. The output header is an optional one line of text which you can enter into the Output Header field of the Page/Output Setup dialog. The text can be used to identify the current project or stage of data analysis and it will appear in the running head of your reports. It will be in effect as long as you do not change it (even after you turn off the computer). It can be automatically included in title areas of graphs and used in other procedures.
    File headers. The file One-line Header (an optional one-line summary title or other identification of the dataset) as well as File Information/Notes (a paragraph of comments) are stored along with the data file and can be entered or edited in the Workbook Info dialog, accessible from the Workbook Info option in the spreadsheet pull-down menu Edit (or by double-clicking on the title field of the spreadsheet displaying the header).
    The one-line header is automatically used in reports and graphs produced from the respective dataset. For example, it is included in printed reports if the selected report style is set to at least the Short setting. The header is also transferred to the title field of custom graphs created from the spreadsheet; it can also be automatically included in the title fields of all graphs created using the Stats Graphs facilities (in the pull-down menu Graphs).
    The File Information/Notes can be reviewed in the dialog shown above (accessible by double-clicking on the title field of the data spreadsheet) and printed, if requested, in the data printing options dialog.

Since STATISTICA uses Julian format for dates, will there be any problems with the dates in the year 2000?
    No, STATISTICA is year 2000 compliant (by design, and confirmed in an extensive set of internal tests). In fact, they are "Year 3000 Compliant" and "Year 4000 Compliant" also.
    In addition to "Year 2000 Compliance", we have worked to enhance handling of date variables in STATISTICA by allowing the user to custom define the default behavior of the program when ambiguous dates are entered. Specifically, as was common in Spreadsheet style packages, STATISTICA by default took a two digit number xx to be 19xx when converting to a year. The 100 year range for which only two digits are required is now under the control of each user. For example, by default the cut-point has been moved to 1920 and therefore typing '00' will be displayed as 2000, '19' will be 2019 while '20' remains 1920. This default cut-point year (1920) can be set to any value between 1900 and 1999, by entering the following line:
        DateCutOffYear=19yy
into the "[Statistica]" section of the STATIST.INI file (where 19yy is the new cut-point year).

When I execute my SCL program, I get the following error message: "SCL syntax error: Declared file does not exist?"
    Before executing the SCL program, STATISTICA will check to make sure files referenced in the program exist (except those specified as part of the SAVEAS command). If the goal of the program is to import and save a file, and then perform various analyses on that file, the file will not exist before the program is executed. To disable the "file checking", go to the Options pull-down menu and de-select "Check Files".

SCROLLSHEETS (Scrollable Tables with Results)

What are Scrollsheets?
    In STATISTICA, all numeric and text output is displayed in spreadsheet-like windows called Scrollsheets. Their content can be anything from a short line to megabytes of output, and they offer a variety of options to facilitate reviewing the results and visualizing them in predefined and custom-defined graphs.

What are the differences between Scrollsheets and spreadsheets?
    All differences between these two--on the surface very similar--main types of "documents" in STATISTICA are related to the difference in their contents: Scrollsheets are used to display results from analyses, whereas spreadsheets hold input data files. Both types of tables look similar and they offer very similar facilities, including the graphics options (both Custom and Stats graphs are available from Scrollsheets and from spreadsheets). Also, both offer easy access to Quick Basic Stats facilities and Block Stats facilities.
    However, each of them offers specialized features and options which are applicable to their respective contents. For example, data spreadsheets feature integrated data management, restructuring, transformation and recoding facilities, whereas Scrollsheets offer specialized graphs related to the types of statistical output they contain or integrated facilities to identify cells or rows containing results which meet particular statistical criteria (e.g., significance level).
    Drag-and-Drop facilities. The standard (ms Excel-style) Drag-and-Drop facilities (moving, copying, clearing, inserting, duplicating, and extrapolating series) are supported in spreadsheets; moving, copying, clearing, duplicating and extrapolating of selected blocks are supported in Scrollsheets.
    Toolbars, flying menus. These differences and similarities are reflected in the appearance of the toolbars which accompany each of the two types of document-windows, as well as in the contents of the flying menus which can be brought up by clicking the right-mouse-button on individual cells.
    Note that these two types of documents can easily be converted into one another (see the next two paragraphs).
    Converting data spreadsheets into Scrollsheets. The Convert to Scrollsheet... option (see the spreadsheet File menu; keyboard shortcut is the F11 key) can be used to convert the data file or the currently highlighted block of data into a Scrollsheet; an intermediate dialog will allow you to select options. A data file saved this way into a Scrollsheet can later be opened (e.g., for reference) at any point of the data analysis without affecting the main data file residing in the data spreadsheet and used as input data (the default file name extension for Scrollsheet files is *.scr).
    Converting Scrollsheets into standard data files (spreadsheets). The Save as Data... option (see the Scrollsheet File menu) can be used to convert the contents of the current Scrollsheet or the currently highlighted block into a standard STATISTICA data file. A data file created this way can later be opened for input as a regular data file, thus allowing you to submit output from one analysis as input for another.

What is the Scrollsheet queue?
    Queues of Scrollsheets (and graphs) Statistical analyses may produce large amounts of output. Scrollsheets offer flexible ways to organize the output regardless of its size. New Scrollsheets are generated by subsequent analyses in a "queue," where older Scrollsheets are closed automatically as new ones are created (to avoid having too many open windows at once). The Scrollsheets are closed on a first-in-first-out basis, and the default length of the queue is 3. In other words, when the fourth Scrollsheet is created, then the first one is closed (with no warning unless you have edited or customized it). The same queue conventions apply to graph windows.
    Note that an option is provided to automatically print (or send to the output file and/or the on-screen Text/output Window) each Scrollsheet which is generated on the screen (see below).
    Auto-report (Automatically Print All Scrollsheets). In order to create a log of all Scrollsheets, select the Automatically Print All Scrollsheets (and optionally graphs) option(s) in the Page/Output Setup dialog (accessible in the pull-down menu File or by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window).
    In the same dialog, you will be able to specify where to direct the output: to the printer, a disk text output file, or the scrollable Text/output Window, and adjust other output settings.
    The length of the queues. In some instances, you may want to increase the length of this queue. Use the Scrollsheet Manager in the pull-down menu Window to temporarily change the length of the queue for the current session (the setting will return to the default length when you close the current module of STATISTICA). In addition, the Scrollsheet Manager dialog (see above) can be used to "lock" specific Scrollsheets (see the next two topics).
    The pull-down menu Options can be used to adjust the queue length permanently (the default length of the Scrollsheet queue is 3 windows):

How to keep a Scrollsheet from being replaced in the queue:
    Regardless of the length of the queue, you can "lock" individual windows (i.e., "remove" them from the queue; use the Scrollsheet Manger accessible from the pull-down menu Window, shown in the previous topic), so that they will not be automatically closed as long as you do not exit the program. Locked Scrollsheets are marked with the @ character appended to the Scrollsheet name in the title bar.

What is the difference between locking and saving Scrollsheets?
    Locking a Scrollsheet (see the previous topic) will preserve the Scrollsheet as long as you do not close the current module. Saving Scrollsheets (see below) allows you to preserve them across analyses (e.g., in different modules or sessions), because saved Scrollsheets can be opened again at any point of the analysis in any STATISTICA module. Also, saved Scrollsheets (file name extension *.scr) can be printed in batch using the Print Files... facility (in the pull-down menu File).

How to save a Scrollsheet:
    Saving Scrollsheets in a Scrollsheet format. Use the option Save (SHIFT+F12) or Save As... (F12) in the File menu. By default, Scrollsheets are saved in the Scrollsheet file format (file name extension *.scr); saved Scrollsheets can be opened again at any point of the analysis in any STATISTICA module. Such Scrollsheet files also can be printed in batch using the Print Files... facility (in the pull-down menu File, and are automatically added to the Workbook list where you can set some of them to Auto-open and then the selected Scrollsheet files will be opened automatically whenever the dataset is opened.
    Saving Scrollsheets as data files. Use the option Save as Data... in the File menu to save a Scrollsheet as a standard STATISTICA data file which can then be used for input in subsequent analyses (file name extension *.sta).
    For information, refer to the respective section on the differences between spreadsheets and Scrollsheets.

How to add rows or columns to the current Scrollsheet:
    Use the options Append Columns... and Append Rows... in the Scrollsheet pull-down menu Edit or from the Columns and Rows toolbar buttons, respectively. The extra space can be used for notes or input pasted from the Clipboard.

How to select (highlight) the entire Scrollsheet:
    Following the standard spreadsheet convention, click in the upper left corner of the Scrollsheet. This shortcut is useful, for example, when you intend to copy the entire file to the Clipboard.
    Note that by clicking twice in the same area, you will de-select (i.e., select and then de-select) the entire Scrollsheet.

How to expand a block in the Scrollsheet:
    Blocks may be selected (1) by drag-selecting (holding down the left-mouse-button) with the mouse, (2) by clicking in one corner of the block to be highlighted, then scrolling to the desired opposite corner (the original cell will remain selected) and pressing the SHIFT key while clicking in that cell, or (3) by holding down the SHIFT key while using the cursor keys on the keyboard.
    To expand a previously-selected block, you may use the SHIFT-cursor key, or scroll the display with the mouse and press SHIFT while clicking in the desired corner.
    To highlight a large block in "split-pane" mode (shown below), click in a cell in one pane, then scroll to display the diagonally opposite corner in another pane and use SHIFT-click to select the block.
    Variable speed of block highlighting. Note that you can control the speed at which you scroll when you extend a block outside the current display window. By moving the cursor a short distance away from the Scrollsheet (or spreadsheet), you can scroll one line at a time when a block is selected, or you can scroll one page at a time by moving the cursor further away from the Scrollsheet (or spreadsheet).

How to edit the contents of a Scrollsheet cell (and not overwrite it):
    Double-click on the cell. To avoid clearing the contents of the cell (at the point when you start entering a correction), first double-click on the cell before you start typing. This will enter the editing mode and will position the cursor within the cell.
    Press F2. Alternatively, you can follow the "old" spreadsheet convention and press the F2 key, which will also enter the edit mode for the currently highlighted cell.

What graphs are available from Scrollsheets?
    There are two general types of graphs available directly from all Scrollsheets: Custom graphs (which allow you to custom-design graphs from the contents of the Scrollsheets to produce user-defined visualizations of results) and Quick Stats Graphs (which are predefined, general-purpose statistical graphs and specialized graphs related to the output presented in the current Scrollsheet). Also, Block Stats Graphs are available in the second level submenu of the flying menu, which produce summary graphs from rows or columns of the currently highlighted block. Refer to differences between Quick Stats Graphs, Stats Graphs and Custom Graphs for more information .

How to make a Scrollsheet from a data file:
    When you are in the data spreadsheet, press the F11 key (or select the Make Scrollsheet... option from the pull-down menu File). The subsequent dialog will allow you to select a subset (by default, all data will be included, unless a block is currently selected). A Scrollsheet containing the selected data will be created. You can save it in a Scrollsheet file (file name extension *.scr) which can be opened at any point without affecting the main data file selected for input in the current analysis. See the respective section for information on differences between spreadsheets and Scrollsheets.

When I ask STATISTICA to print out a correlation matrix larger than one page, I only get one page of the printout. What is the problem?
    What you receive as a printout is determined in part by what options are selected in the Page/Output Setup dialog. In this instance, if you have selected the Automatically Eject Each Filled Page option, STATISTICA will only print out pages when they are filled with information, and therefore if you are printing out a large correlation matrix, you may not receive "the rest" of this matrix until more output is directed to the printer. A way to fix this problem is to de-select this option and instead select Automatically Print/Eject Pages in the Page/Output Setupdialog.

How to print text/graphics reports from analyses:
    Combined text/graphics reports can be created in the Text/output Window by selecting Window in the Page/Output Setup dialog and pressing the Print button on the toolbar to add individual Scrollsheets and graphs to the report. You may also select the options Automatically Print All Scrollsheets and Automatically Print All Graphs to Output Window to create an automatic log of all output created during an analysis.

How to print Scrollsheets:
    The simplest way to print a Scrollsheet is to press the Print button on the toolbar. If an output channel is not selected, then STATISTICA will open the Page/Output Setup dialog and ask you where to send the output (to the printer, disk file, and/or the Text/output Window). No other intermediate option dialogs will be displayed. The entire Scrollsheet will be sent to the output. If a block is selected in the Scrollsheet, then only that block will be sent to the output destination.
    More options are available when you use the Print... (or CTRL+P) option in the pull-down menu File. The Print Scrollsheet options dialog will appear allowing you to customize various aspects of the Scrollsheet printing. For example, you may print only cells highlighted in the Scrollsheet (leaving other cells blank) in order to emphasize patterns of significant relations in a large correlation matrix.
    Automatic reports. Note that if you intend to keep a complete log of all Scrollsheets (and/or graphs) which are displayed on the screen without having to remember to print them, select the respective Auto-report option in the Page/Output Setup dialog, accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window.
    Other options. In the same Page/Output Setup dialog, you will be able to specify where to direct the output (printer, disk-output file, and/or the scrollable Text/output Window).

How to print multiple Scrollsheets previously saved in Scrollsheet files:
    In order to print a set of previously-saved Scrollsheets (and/or graphs) in batch, use the Print Files option, available from the pull-down menu File, to open the Print Graph and Scrollsheet Files dialog. Use the Add Files button in this dialog (see below) to select the desired Scrollsheets (and/or graphs), and click OK to begin the batch printing process. All of the selected graphs will be printed first (on separate pages, in the order in which they are selected), and then all of the Scrollsheets will be printed together (in the order in which they are selected).

How to print data files:
    The simplest way to print a data spreadsheet is to press the Print button on the toolbar when the spreadsheet window is active. If an output channel is not selected, then STATISTICA will open the Page/Output Setup dialog and ask you where to send the output (the printer, disk file, the Text/output Window). No other intermediate option dialogs will be displayed. The entire spreadsheet is sent to the output. If a block is selected in the spreadsheet, then only that block will be sent to the output. More data printing options are available when you use the Print... (or CTRL+P) option in the pull-down menu File. The Print Data options dialog will appear allowing you to customize the data listing report.
    You can specify the subset to be printed (by default, the current block, if one is selected), the types of data and the categories of information about variables to be included, the text of the Notes attached to the file, and formatting. Other aspects of the report format, including headers, page numbers, margins, etc., can be customized in the Page/Output Setup dialog (accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window).

Can I combine text tables and graphs in one report?
    Yes. Reports combining text and graphics can be created in the Text/output Window by selecting Window in the Page/Output Setup dialog and pressing the Print button on the toolbar to add individual Scrollsheets and graphs to the report.

How to adjust margins in the text reports:
    Use the Text Output Margins dialog (accessible by pressing the Margins button in the Page/Output Setup dialog (accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window).
    You can position the desired text (e.g., from spreadsheets, Scrollsheets, or the Text/output Window) on the printed page by adjusting the margins (specify the Left, Top, Right, and Bottom margin widths) according to a specific measure (Percent, Centimeter, or Inch).
    Note that these margin settings are independent from the margin settings for the graphics output (which can be adjusted in the Print Preview or Graph Margins dialogs).

How to suppress the printing of gridlines in tables:
    The list box Lines in the Page/Output Setup dialog (accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window) allows you to control the thickness of lines printed in tables. Selecting the top setting in the list (the empty field) will suppress the lines (see also the next topic).

How to change the character-lines into continuous lines in report tables:
    Tables printed in STATISTICA. If gridlines of any width are requested in reports, the lines are always printed in STATISTICA as continuous high-resolution lines and not character lines (see the Page/Output Setup dialog, accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window). The character lines will be produced only if requested in the Text/output Window or text file output, where high-resolution lines are not drawn (to allow that output to be edited with all word processors and text editors).
    STATISTICA tables printed in word processors. If instead of printing the reports in STATISTICA (featuring high-resolution gridlines in tables), you need to include high-resolution tables in your word processor documents, it is recommended to use the Clipboard to copy the desired Scrollsheets and spreadsheets and paste them into word processor documents. If pasted this way, STATISTICA Scrollsheets and spreadsheets will appear as tab-delimited tables, compatible with table generators in Windows word processors (e.g., ms Word). Note that numbers in tables in the Text/output Window can be delimited with either spaces or tabs; tabs should be used if the contents of the Text/output Window will be later accessed by a word processor.

How to add custom headers to printed output:
    One line of text can be added to the fixed report header (including the name of the module, page number, and the optional time and date). Enter the text to be used as the header into the Output Header field in the Page/Output Setup dialog (accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window).
    The text will be printed in the second line from the top of every page of the report. This text will also be automatically saved by STATISTICA even when you turn off the computer and it will be included in reports as long as you do not change it.

How to automatically print (or save to report files) all Scrollsheets and/or all graphs from an analysis:
    The Auto-report option. Enable the respective Auto-report switches in the respective Page/Output Setup dialogs (accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window). These modes will automatically generate a log of all Scrollsheets and/or graphs displayed on the screen, releasing you from having to press the Printer button in the toolbar in order to print individual Scrollsheets and/or graphs.
    The contents of every Scrollsheet and/or graph generated on the screen are automatically sent to the current output channel (e.g., the Text/output Window) following the report style and conventions defined in that dialog.
    "Automatic pressing" of the Continue button to speed up batch output. When, instead of reviewing the output on-screen, you need to quickly send it to the report (Text/output) window or produce a hard copy, then it is advantageous to select the Auto-Exit from Scrollsheets and Graphs option in the Page/Output Setup dialog. If that option is selected, STATISTICA will "internally" press the Continue button on every graph and Scrollsheet, thus allowing you to print long sequences of Scrollsheets without having to press the Continue button at the end of every "queue-full" of Scrollsheets.
    For example, if an analysis produces 10 Scrollsheets and the current Scrollsheet queue is set to 3 (default), then STATISTICA will stop after displaying every 3 Scrollsheets and wait for you to press the Continue button, regardless of whether the Scrollsheets are automatically sent to the output or not. However, when the Auto-Exit from Scrollsheets and Graphs option is selected, then all 10 Scrollsheets will be displayed (and printed) one after another without prompting you for any input.

How to speed up printing reports:
    Turn off the option Automatically Eject Each Page. This option is included in the Page/Output Setup dialog (accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window). When the Automatically Eject Each Page option is enabled (a default setting), then each page of the printout is finalized as soon as it is full. Thus, this mode is somewhat slower overall, as the process of printing is initiated for each page separately; however, its advantage is that it automatically produces hard copies as your analysis progresses (which may be desired in some circumstances). Alternatively, when this switch is turned off, all output is accumulated in an internal buffer and printed more economically all at once at the end of the session (or whenever you select the Print/Eject Current Pages option in the pull-down menu File, which allows you to "flush out" the current contents of the output buffer at any point).
    Do not request the high-resolution table gridlines. Note that table gridlines are printed as high-resolution graphics, thus the output is faster when the gridlines are suppressed (set the Lines field to the empty field setting).
    Adjust the report style. If the report style is set to Medium or Long (in the Page/Output Setup dialog, accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window), then large amounts of supplementary information about the selected variables are automatically directed to the printer (press the Help button or the F1 key in that dialog to review descriptions of the report styles). The amount of this supplementary information is particularly large when the variables have many text values and long text value labels. Changing to a more economical style (e.g., Short) will speed up printing.

Can I append output from multiple sessions to the same Text/output Window?
    Yes. Check the Auto-Retrieve the Text/output Window Contents option in the Page/Output Setup dialog. New output will then be added to the previously generated reports.

How is the disk file output different from the printed output?
    If the output is set to FILE (in the Page/Output Setup dialog, accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window) then STATISTICA will ask you for the name of an output file, and then direct to that file a text representation of all output. The text contents of the reports directed to the output file are identical to those which are printed (if the output is set to PRINTER), with one difference mentioned below.
    Gridlines in tables. The only difference between the printer and file output of text is that unlike in the printed output, character (instead of continuous) table gridlines are produced in disk file reports to assure their compatibility with all word processors and text editors. (Click here for information on how to print STATISTICA tables with high-resolution gridlines in word processors.)
    Continuous reports from different modules. The output continues to the same output file as you switch modules of STATISTICA. Moreover, if you do not turn off the output before exiting the program, next time when you start STATISTICA, it will automatically open the last used output file and continue appending the new output to it.
    Changing output files. You can close the output file and start a new one at any point via the Page/Output Setup dialog (accessible by double-clicking on the Output field in the status bar on the bottom of the STATISTICA window). Press the Output file button and enter the new output file name.
    When the output file cannot be edited. The Text/output Window contents can always be edited, however, the output file cannot be edited (e.g., cannot be opened in the STATISTICA text editor) at the same time that it is open for output. You can edit it only after it is closed. If you wish to edit that file, close it first by turning off the file output or selecting a new file for output (see the previous paragraph).

What is the difference between the output to the on-screen Text/output Window and a disk-output file?
    These two output channels can be open simultaneously, both have practically unlimited capacity, and they receive the same output (selectable in the Page/Output Setup dialog, accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window). However, they differ in several ways.
    Disk-output files. The disk file output produces permanent report files containing no special formatting characters and no representations of graphs. The reports can be edited using STATISTICA or any other text editor or word processor, but not when the file is open by STATISTICA for output (see the last paragraph of the previous topic).
    On-screen Text/output Window. On the other hand, the on-screen Text/output Window is like a graphics/text editor which is open simultaneously for input from several sources: it receives STATISTICA's "printed" text output (Scrollsheets, spreadsheets) and graphs, and at the same time you can add to it your own notes, or edit and supplement it by pasting from the Clipboard text (including formatted, RTF-compliant text), and custom graphs or other graphics objects. The Text/output Window can be used as an editable, temporary means to keep a log of and review recent output, or as a facility to generate text/graphics reports (of practically unlimited size) to be printed from STATISTICA or transferred (via Clipboard or *.rtf files) to a word processor or desktop publisher for further editing.

The vertical gridlines do not show properly in printed tables; how do I change the printer setup?
    Enable the Print TrueType as Graphics option in the Setup/Options dialog of Printer Setup... (pull-down menu File). The setup programs for some drivers contain this switch in the Options... dialog.

GRAPHICS

General

What categories of graphs are available in STATISTICA? What is the difference between Custom graphs and Stats graphs?
    In addition to the specialized graphs which are available from the output dialogs of all statistical procedures, there are two general categories of graphs accessible from menus and toolbars of all Scrollsheets and spreadsheets:

• Custom Graphs, and
• Stats Graphs (and Quick Stats Graphs).

• Quick Stats Graphs can be called from flying menus and toolbars (and they do not require any pull-down menu selections),
• Quick Stats Graphs do not require the user to select variables (the variable selection is determined by the current cursor position within a Scrollsheet or spreadsheet), and
• Quick Stats Graphs do not require the user to select options from any intermediate dialogs (default formats of the respective graphs are produced).

Are there different types of customization options available for each of the types of graphs described in the previous topic?
    No. Regardless of how a graph was requested or defined, once a graph appears on the screen, all graph customization options available in STATISTICA (see the topic How to customize graphs, below) can be used to customize it. The customization options available for all graphs also include appending new plots to existing graphs, merging graphs, linking and embedding graphs and artwork, as well as all drawing, fitting, and graph restructuring options. Also, all those options can be used to customize graphs which were saved and later retrieved from disk files.

How to access graphs which visualize raw data:
    This topic is a brief summary of the first topic of this section (What categories of graphs are available in STATISTICA?). Note that all three types of graphs listed below maintain automatic links to data from which they were created and (unless the manual update is requested or the link disturbed) they are updated automatically when the data change.
    As mentioned previously, there are several "categories" of graphs available in STATISTICA which visualize raw data (as opposed to visualizations of results of analyses or other graphs based on pre-processed data).
    (1) Custom Graphs. When requested from the spreadsheet (via toolbars or flying menus), Custom Graphs allow you to custom-define any aspects of raw data (cases, variables or both) and represent them in graphs. While selecting these graphs, the user is not confined to any pre-defined formats or layouts.
    For example, you can request a scatterplot of data in Case 1 against data in Case 145 taking into account only variables number 17 through 225; or you can request a 3D XYZ plot of the values of some cases against the values of some variables (see the first topic in this section for details).
    (2) Quick Stats Graphs. Quick Stats Graphs offer the quickest method to create the most-commonly used (predefined) statistical graphs from spreadsheets or Scrollsheets (e.g., histograms, scatterplots, box and whisker plots, probability plots, etc.).
    These graphs are available either from the toolbars or flying menus, and because the selection of variables to be plotted in these graphs is determined by the current cursor location (within the current spreadsheet or Scrollsheet), Quick Stats Graphs require a minimum of input on the part of the user. Most of these graphs require just two clicks of the mouse.
    (3) Stats Graphs. As opposed to Quick Stats Graphs (see the previous paragraph) which are designed to offer the quickest access to the most-commonly used types of statistical graphs (from raw data), Stats Graphs (available from the pull-down menu Graphs) offer a large variety of graphs and graphical analytic procedures selectable via a user-interface which follows the logic used to specify all statistical or analytic procedures in STATISTICA. This logic is illustrated in the next topic.

How to specify Stats Graphs: Example of a complex, categorized graph
    As mentioned in the previous topic, STATISTICA's Stats Graphs (which are a "superset" of Quick Stats Graphs) offer a large variety of graphical analytic procedures selectable via a user-interface which follows the logic used to specify all statistical or analytic procedures in the STATISTICA system.
    Customization of Stats Graphs before vs. after a graph is produced. As described in the section How to customize STATISTICA graphs), most graph customization options (hundreds of settings which control virtually all aspects of the appearance of STATISTICA graphs) are interactively accessible after a graph is produced, by clicking on specific components of the graph (or from the graph General Layout and Plot Layout dialogs accessible from the pull-down menu Layouts).
    However, some of the graph definition or customization features--specifically, those which control how data are accessed from the data file, and what information is extracted from those data (e.g., variables, categorization, value labels, case names, scale value labels)--have to be set before the graph is produced. In this example, the selection of variables, the categorization of cases, and the settings available in the dialog accessible by pressing the Options button (not used in this example), had to be set before the graph was produced.

Can graphs be automatically updated when the data file changes?
    Yes, all Stats, Quick Stats, Block Stats, and Custom Graphs produced from raw data (spreadsheets) maintain automatic links to the data from which they were created and (unless the manual update is requested or the link disabled) they are updated automatically when the data change. Links can be managed in the Graph Data Links dialog (accessible from the pull-down menu Graphs).
    Links may be automatic (updating the graph whenever the data from which it was created are changed), changed to Manual, or temporarily locked. Also, the links (which are by default set up for the specific data file from which the graph was created) can be set to the Current Data File mode, allowing you to recreate the same graph or sets of graphs for different data files. The linking mechanism can be configured permanently from the Options pull-down menu.
    Links can also be "nested" and start in another application. For example, an Excel 5 spreadsheet can be connected to a STATISTICA dataset via DDE and the data can be linked to graphs: When you press the F9 key in Excel to recalculate (e.g., randomize), STATISTICA will update automatically and this will update all graphs. See also the next two topics.

Can STATISTICA automatically open graphs related to a data file and update them automatically when the data change?
    Yes. Graphs associated with a dataset can be marked in the Workbook as Auto-open, so that whenever the dataset is opened, the graphs so marked are opened automatically. The graphs will be updated to reflect the current data values when you make a change to a relevant part of the data file.

Can I automatically recreate a set of specific graphs for a new data set?
    Yes. To allow for these types of operations, the links between data in graphs and the data file from which they were created are maintained in terms of variable and case numbers (not names), and they can be linked to the Current Data File (instead of the default setting: the specific data file from which the graph was produced; see the Graph Data Links dialog, above). Thus, after creating (or opening from disk files) a set of graphs, you can simply open another data file and all graphs will be automatically recreated for the data residing in the respective column numbers of the new data file.
    For example, a list of files to be compared via the same set of graphs can be included in the File menu (under the list of recently used data files); now you can quickly open those files one after another and watch changes to all graphs which will automatically update.
    Also, both the STATISTICA BASIC and SCL languages can recreate user-specified sets of graphs for different datasets. Note that a set of STATISTICA BASIC or SCL programs (or a sequence of keystrokes) that specify a series of graphs can be assigned to an Auto Task Button. Also, previously-saved graphs can be opened by assigning their names (directly) to Auto Task Buttons and those graphs will automatically update (upon opening) unless you disable their links to the data.

What is the Graph queue?
    The queues of graphs. Statistical analyses sometimes produce large numbers of graphs (e.g., cascades of graphs for each combination of variables from two lists). Queues of document windows implemented in STATISTICA offer flexible ways to organize such output regardless of its size. New graphs are generated by subsequent analyses in a "queue," where older graphs are closed automatically as new ones are created (to avoid having too many open windows). The graphs are closed on the first-in-first-out basis, and the default length of the queue is 3. In other words, when the fourth graph is created, then the first one is closed (with no warning unless you have edited or customized it). The same queue conventions apply to Scrollsheet windows. Note that an option is provided to automatically print each graph which is generated on the screen (see below).
    Automatically printing all graphs. In order to create a log of all graphs, select one of the Auto-report options in the Page/Output Setup dialog (accessible in the pull-down menu File or by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window). For example, you can print all graphs or send them to the Text/output Window. For more information about this option, refer to the section on Printing Graphs.
    The length of the queue. In some instances, you may want to increase or decrease the length of this queue. Use the Graph Manager in the pull-down menu Window to change the length of the queue for the current session (the pull-down menu Options can be used to adjust the queue length permanently).

How to keep a graph from being replaced in the queue:
    Regardless of the length of the queue, you can also "lock" individual windows (i.e., "remove" them from the queue; use the Graph Manger in the pull-down menu Window), so that they will not be automatically closed as long as you do not exit the program. Locked graphs are indicated with a @ character appended to the graph title in the title bar.

What is the difference between locking and saving graphs?
    Locking a graph (see above) will preserve the graph as long as you do not close the current module. Saving the graph (see below) allows you to preserve it across analyses (e.g., in different modules) or open it again in a future session.

How to save STATISTICA graphs:
    STATISTICA graphics file format. Graphs and drawings may be saved in the STATISTICA system graphics file format (file name extension *.stg). Use the options Save or Save As... in the File menu. Although other graphics formats are also supported, this format is recommended whenever you intend to use the graph again in the STATISTICA system or link it to other application documents using the Windows OLE conventions. Unlike the other formats (see the next paragraph), the STATISTICA system format stores not only the graphical representation of the picture but also the Graph Data Editor (see below) containing all data which are represented in the graph, all analytic options (fitted equations, ellipses, etc.) and other settings allowing you to continue the graphical data analysis at a later time. This format is also most appropriate if the current graph or drawing is later to be linked to or embedded into another STATISTICA graph. Graphs saved in this format can also be printed in batch using the Print Files... facility in the pull-down menu File.
    Bitmap and Metafile graphics format. If the graph to be saved is to be used by an application which does not support OLE, use the Save Metafile... or Save Bitmap... options (in the File menu) to store the current graph in the Windows metafile (file name extension *.wmf) or Windows device-independent bitmap format (file name extension *.bmp), respectively. Those formats (described briefly in the next two topics) do not offer the advantages of customizability offered by the STATISTICA format (see above), however, they are compatible with all applications which support Windows graphics file formats.

What is the Windows Metafile graphics format?
    The Metafile format, also referred to as Picture (one of two standard Windows graphics formats used in Clipboard and disk file representations of graphs; file name extension *.wmf) stores a picture as a set of descriptions or definitions of all components of the graph and their attributes (e.g., segments of lines, colors and patterns of those lines, specific fill patterns, text and text attributes, etc.). Therefore, as compared to bitmaps (another standard Windows graphics format, see below), metafiles offer more flexible options for non-OLE modification and customization in other Windows applications.
    For example, when you open a metafile in the ms Draw program, you can "disassemble" the graph, select and modify individual lines, fill patterns, colors, edit text and change its attributes, etc. Note, however, that not all Windows applications support the complete set of metafile graphics features and attributes supported and used by STATISTICA, thus some aspects of STATISTICA graphs (saved as metafiles) will look different when they are opened in such applications (e.g., rotated and transformed fonts may appear unrotated). For best results, copy STATISTICA graphs to other applications using the OLE conventions which allow you to invoke STATISTICA to customize or modify STATISTICA graphs pasted into documents of other applications.
    Technical Note: Limitations of the Metafile format. The Windows graphics Metafile format is well-suited for representing small to moderate size (i.e., complexity) graphics images. However, when the file size exceeds 1-2 megabytes (depending on its contents) the Windows system may interpret the graph incorrectly (e.g., the Clipboard mechanisms may malfunction or the graph file may be distorted). STATISTICA will beep or issue an error message when the metafile to be created would exceed the Metafile format limitations (see the Advanced options in the Graphs layer of the Page/Output Setup dialog). STATISTICA may generate graphs of size and complexity which exceed by far the capacity of the metafile representation; if you need to transfer such graphs to other applications, use the Screen Resolution option for Metafile Mode (in the Page/Output Setup dialog), which will produce smaller metafile representations (both in Clipboard and when saving to the metafile format), or use the Bitmap representation (either via the Clipboard - use Paste Special in the target application, or disk files - use the Save Bitmap option) which does not impose any file size limitations.

What is the Bitmap graphics format?
    The Bitmap format (one of two standard Windows graphics formats used in Clipboard and disk file representations of graphs; file name extension *.bmp) is similar to another standard Windows graph format--metafile (see above), in that it stores only the representation of the picture (and not the data which are plotted or any information about the analytic properties or settings used to produce the graph).
    However, unlike the metafile format, the bitmap format stores only a "passive" representation of pixels which form the graph. This representation is thus less customizable than Metafiles which store dynamic representations of all individual graph components, thus allowing selective modifications of lines, text, etc., in other Windows applications.
    Bitmaps can be opened by other Windows applications, but the customization or editing options of such graph representations will be limited (typically to operations on pixels, such as stretching and shrinking, cutting and pasting, and drawing "over" the graph). As mentioned before, for best results, copy STATISTICA graphs to other applications using the OLE conventions which allow you to invoke STATISTICA to customize or modify STATISTICA graphs pasted into documents of other applications.

What is the Native STATISTICA graphics format?
    Unlike other graphics formats such as metafiles and bitmaps, the native STATISTICA graphics file format (file name extension *.stg) not only stores the "image" of the current graph but also all information necessary to continue graph customization or graphical data analysis (including all data represented in the graph, links to data, fitted equations, embedded graphs and artwork, links to graphs and artwork, etc.). Graphs stored in this format can be opened later (in any module of STATISTICA) allowing you to continue the graphical data analysis or graph customization (e.g., fit equations, add or merge new data series, etc.). Graphs saved in this format can also be printed in batch using the Print Files... facility in the pull-down menu File. They can also be dynamically linked to documents of other Windows applications using the OLE conventions.

How to export a STATISTICA graph to another application:
    Export via the Clipboard (and Paste or Paste Special via OLE). The quickest way to export a graph is to copy it to the Clipboard and then paste it into another application. STATISTICA native, Windows metafile, and bitmap formats are created in the Clipboard and can be used in other applications.
    STATISTICA graphs can be pasted to other application documents (e.g., word processor documents, spreadsheets) either as embedded objects or as objects linked to graph files. If STATISTICA graphs are pasted to other applications via Windows OLE, they maintain their relations to STATISTICA and thus can be interactively edited from within the other application.
    Linking STATISTICA graph files via OLE. STATISTICA graph files can also be inserted and linked via OLE to other applications and they will maintain their dynamic relations to the STATISTICA source files and thus they can be automatically updated when the STATISTICA source files change.
    Export to another file format. STATISTICA graphs can also be saved to other graphics file formats; use the Save Metafile... or Save Bitmap... options (in the File menu) to store the current graph in the Windows metafile (file name extension *.wmf) or Windows device-independent bitmap format (file name extension *.bmp), respectively.
    Limitations of the standard Windows Metafile format. Very large (in terms of the number of data points represented) or very complex graphs that can be produced by STATISTICA may exceed the capacity of the Windows Metafile graphics format used in the Windows system as a default graphics representation (e.g., in most OLE operations); in those circumstances you should use the Bitmap representation instead (see Limitations of the Metafile format). For more information, refer to the Electronic Manual from the Advanced Options dialog, accessible from the Graphs layer of the Page/Output Setup dialog.

How to stop drawing or redrawing a graph:
    Press any key or click the mouse anywhere on the screen. If the redrawing is related to some smoothing or function-fitting calculations and the progress bar is displayed, then press the Cancel button on the progress bar (see the next topic).

How to stop function-fitting calculations:
    Press any key, click the mouse anywhere on the screen, or (if the progress bar with the timer is displayed on the status bar) press the Cancel button located at the end of the progress bar. If the interruption of processing will involve losing some calculations that have already been performed, STATISTICA will ask you for confirmation before the processing is halted.
    If you answer No, then the processing will resume. Note that STATISTICA supports multitasking, thus you can switch to some other application (including other modules of STATISTICA) and let the time-consuming process be completed in the background.

How to resume an interrupted redrawing of a graph:
    Redrawing of an interrupted ("frozen") graph can be resumed either by pressing OK in any of the graph customization dialogs (see the pull-down menu Layouts or click the right-mouse-button anywhere on the graph) or by touching the graph frame with the mouse as if you intended to re-size the graph window.
    Selecting the option Restore Original Settings in the graph pull-down menu View will have the same effect but the graph will be redrawn in its default graph-window arrangement (i.e., using the default proportions, size in the window, and margins settings).

What is the difference between a graph and a plot?
    Each plot represents a single "series" of data. All but the most simple graphs in STATISTICA contain more than one plot of data.
    In other words, in STATISTICA, the term graph applies to a complete graphical representation of (one or more) "series" of data, that is, to the entire "picture," which can be saved as a graphics document (by default, as a STATISTICA graphics file, file name extension *.stg).
    There are many ways in which plots can be "put together" to form a graph, and depending on their type, some plots may require more than one sequence of values.
    For example, at least three coordinated sequences of values are necessary to create a whisker plot: X-values, Y1-values (lower endpoints of whiskers), and Y2-values (upper endpoints of whiskers).
    Plots in the Graph Data Editor. One Graph Data Editor window (shown above, see also the next topic for an overview) contains all data represented in a graph. In other words, it contains data for all plots in a single graph, and each plot is represented in the Editor by one column (typically a double- or triple-column).
    The same general principle of "graphs consisting of plots," where each plot is represented by a custom-formatted column in the Graph Data Editor (see the next paragraph) applies to all types of graphs in STATISTICA (2D, 3D, icon, and matrix).
    Customizing the layout of a graph (the General Layout dialog). The layout of a graph contains all those features and attributes which apply to the entire graph and are common to all plots. They include such features as titles, gridlines, global colors (backgrounds, etc.), scaling, axis labels, or categorization labels. They can be adjusted in the General Layout dialog accessible from the pull-down menu Layouts, or via flying menus (called by clicking the right-mouse-button on the graph or the Graph Data Editor, see below).
    Customizing the layout of a plot (the Plot Layout dialog). On the other hand, the layout of a plot contains patterns, sizes, and all other specifications which apply to the graphical representation of only one series of data. They can be customized in the Plot Layout dialog accessible from the pull-down menu Layouts, via flying menus (called by clicking the right-mouse-button on the graph or the Graph Data Editor), or by clicking on the respective column of the Graph Data Editor (see the next topic).

What is the Graph Data Editor?
    Access to all graph data. In STATISTICA, all values represented in every graph can be reviewed and edited directly. In other words, regardless of whether a graph represents raw data from the data spreadsheet, parts of a Scrollsheet output, or a set of calculated or derived scores (e.g., in a probability plot), these values are always accessible along with the graph via the internal Graph Data Editor.
    There is one such Editor associated with each graph and managed as a "child window" of the graph (the Editor window will close when the graph is closed). The Editor is organized into column-segments representing individual plots (i.e., series of data, see the next paragraph) from the current graph.
    Columns representing individual plots. In mixed-type graphs, each column-segment may represent a different type of plot (e.g., line plot, scatterplot), and those respective types are marked by icons in the column name areas of the Editor. The column-segments may consist of single-, double-, triple-, or quadruple-columns of values (depending on the type of the respective graph, see the previous topic). The legend of each plot is displayed in the header of the respective column (next to the icon representing the type of plot). Double-clicking on that column header will bring up the Plot Layout dialog, allowing you to edit the legend as well as adjust all other features and characteristics of the plot (patterns, fitted functions, etc.).
    Data management options in the Graph Data Editor. In addition to the standard Clipboard-based and Drag-and-Drop-based editing, global restructuring operations can be performed on graph data. The contents of the Editor can be expanded (see the options Add Plot and Add Rows in the Editor pull-down menu Edit, or the flying menu accessible by clicking the right-mouse-button on the Graph Data Editor), combined with other data, or saved in tab-delimited data files, etc. (for more information, press the F1 key in the Editor). When you save the graph, the complete contents of the Editor are also stored in the graphics file, so that later you can continue interactive data analyses (e.g., brushing).
    Data point labels. Data point labels can be edited in the first column by double-clicking on the label (the Graph Data Point ID Info dialog will open, see below); if different plots have different labels, then the contents of the first column will change, depending on the cursor position, so that the labels for the currently highlighted plot are displayed.
    Attributes of individual data points. In addition to labels, individual data points in the graph may have other attributes, and those can be reviewed in the Graph Data Point ID Info dialog (shown below), accessed (1) by pressing the Graph Point ID button in the Graph Data Editor toolbar, (2) by selecting Point Label/ID from the Edit pull-down menu, or (3) from the right-mouse-button flying menu in the Graph Data Editor. In this dialog, the label, subgroup membership (by category or subset) and attributes of selected points are displayed and can be changed (see Selected data points in the Graph Data Editor, below, for details on points selected by brushing).
    Toolbar in the Graph Data Editor. The toolbar buttons in the Graph Data Editor offer the standard data formatting and editing facilities. In addition, the toolbar includes the Graph Data Point Label/ID button, two buttons which initiate redrawing of the current graph: Redraw and Exit and Redraw (the latter closes the Editor), and two buttons providing quick access to the two main graph customization dialogs: Plot Layout, and General Layout. The toolbar also includes five buttons which control the attributes of designated points (see below for a glossary of brushing terms). When you press any of these buttons, the respective attribute will be assigned to the data point in the current cell (or block of cells):

• Marks data points,
• Labels data points,
• Turns off (i.e., temporarily eliminates) data points,
• Highlights data points (this attribute is visible only when the graph is in brushing mode),
• De-selects data points (i.e., removes all attributes).

What are categorized graphs?
    Categorized graphs are created by categorizing data into subsets and then displaying each of these subsets in a separate small component graph arranged in one display. For example, one graph may represent male subjects and one female subjects, or high blood pressure females, low blood pressure females, high blood pressure males, etc.
    In STATISTICA, categorized graphs are:

• available in many output dialogs (they are automatically generated as part of output from all procedures which analyze groups or subsets of data, e.g., breakdowns, t-tests, ANOVA, discriminant function analysis, nonparametrics, and many others),
• accessible as part of the Quick Stats Graphs options in the flying menus in all Scrollsheets and spreadsheets, and
• accessible as part of Stats Graphs (from the pull-down menu Graphs) where a wide variety of user-defined methods to categorize data are available.

How to define "categories" for categorized graphs:
    If categorized graphs are requested from output dialogs of specific procedures which involve analyses of subsets of data, then they will automatically display the subsets which are currently analyzed (i.e., the subsets are already defined as part of the current analysis). On the other hand, the categorized Stats Graphs requested from the pull-down menu Graphs offer a variety of methods to specify subsets using one or two grouping variables. Also, custom-defined subset definitions can be used which can involve all variables in the current dataset.
    Specifically, categories can be defined by:

• Integer values of grouping variables (Integer Mode),
• Dividing grouping variables into a requested number of equal-length intervals (Categories:),
• Custom intervals (ranges) of grouping variables, defined by specific interval boundaries (Boundaries:),
• Specific values (i.e., codes) of grouping variables (Codes:), and
• User-defined "multiple-subset" definitions (Multiple Subsets) which can be entered as logical case selection conditions of virtually unlimited complexity (this categorization method may involve values of all variables in the current data file).

How to display definitions of subsets which are too long to fit in the Fixed legend:
    The length of fixed legends which display subset definitions is limited to 24 characters; if your subset definitions are longer, then instruct STATISTICA to place the text of definitions in a movable legend (which has the same status as custom text entered into the graph and thus its length is not limited).
    In the graph definition dialog for the respective graph (the one from which you select the variables to be displayed in the graph), access the Stats Graphs: Options dialog by pressing the Options button (this button is always located in the upper part of the dialog, below the buttons OK and Cancel). The Stats Graphs: Options dialog contains global settings which are common to most Stats Graphs. Select the option Create long legends for mult. subset graphs; the legend containing the subset definitions will now be created as custom text (of practically unlimited length).
    The initial position of such long legends may not be optimal, however, you can move it, reformat, reduce, place in fixed titles, etc. Note that all fixed legends can also be converted into movable legends, however, the length of text which can be placed in fixed legends (at the point when the graph is created) is limited, therefore, if your subset definitions are long, it is recommended to create the initial legend as moveable.

How to produce ternary contour plots and surfaces:
    Ternary plots of 3D surfaces or contours can be produced as part of the output from the analysis of mixture designs in the Experimental Design module.
    Ternary plots are also available from the Stats 3D XYZ Graphs, Stats Categorized Graphs, and Custom Graphs options on the Graphs pull-down menu.
    For more information, please refer to the Electronic Manual (accessible by pressing the Help button or the F1 key).

How to produce graphs with polar coordinates:
    Polar coordinates may be selected in the graph definition dialogs for scatterplots, line plots, and stacked/sequential plots from the Stats 2D Graphs menu (accessible from the pull-down menu Graphs). Categorized graphs with polar coordinates can also be produced.
    In addition, many standard (Cartesian) graphs may be converted into polar coordinates by switching to Polar in the General Layout dialog. For more information, please refer to the Electronic Manual (accessible by pressing the Help button or the F1 key).

How do I make a line plot of x-y values and not just sequential values versus casename (or case number)?
    Choose the Stats 2D Graphs and then 2D Line Plots (Variables) from the Graphs pull-down menu. Select XY Trace under the Graph Type section and then press the Variables button and select the variables representing x and y. Click OK to accept the variables and then click OK again to create the graph.

Why does the title in my scatterplot not update when I change the fit being used?
    This happens only when in the title there is no formula linked to the data (but only a part of statistical output from the previously requested analysis). Instead of making the graph from the Correlations dialog (or other output control dialogs), a Scatterplot from the Stats Graph option (available from the Graphs pull-down menu or the Graphs Gallery toolbar button) should be used. With Stats Graphs, when the fit, scaling, and/or brushed points are changed, then the fit is automatically recalculated and placed in the title area.

Currently, when I make a graph, STATISTICA makes the graph and then automatically sends me back to the graph selection dialog, without letting me look at the graph. How can I fix the program, so that I can see the graph when it is made?
    This happens when the "Auto exit" option is selected. In the Page/output Setup dialog (available by double-clicking in the Output box in the bottom right hand corner of the STATISTICA application window), de-select the option to Auto exit from Scrollsheets and Graphs.

When I have set up a graph using selection criteria and saved that criteria to a *.sel file, why does STATISTICA not automatically use that criteria when updating the graph?
    When the data file is closed, notice that the SEL option in the Status Bar at the bottom of the screen changes from ON to OFF. There is no way to directly have the Selection Conditions file (*.sel) open up automatically when you open the data file (*.sta). However, after opening the data file, you can activate your selection criteria by double clicking on the SEL option in the Status Bar at the bottom of the screen, clicking the Open button, selecting the file with your selection criteria, and clicking Open. Make sure the ON radio button is selected under the Status section. Click OK. Note that the SEL option in the Status Bar at the bottom of the screen changes from OFF to ON. Now, open your graph file and the criteria should be used when updating the graph. You can use a macro to automate this task.

How can I collapse groups or values when creating a histogram or pie chart?
    You may use the categorization method by selecting the Boundaries radio button in the Stats 2D Graphs dialog instead of the default method. You may then select the Specify Boundaries button and enter the upper cutoff values for each respective bin, separated by spaces.

Graph Customization - General Features

How to customize STATISTICA graphs:
    Customization before vs. after displaying a default graph. All graph customization facilities in STATISTICA are available when a graph window is active (thus, after a graph has been shown). Usually, graphs appear on the screen instantaneously, and typically, it is preferable to first see the default appearance of the graph before proceeding with any customizations. However, in the rare cases when producing a graph takes more time (e.g., in case of complex multi-graphs or graphs based on very large datasets), you can also make the customization before the drawing of the default graph is completed. Specifically, you can always interrupt the drawing with a single click (press any key or click the mouse anywhere on the screen). Then, you can access some customization tools or facilities and resume redrawing after making the desired adjustments.
    There are two major types of graph customizations:

• adding/editing custom graphic objects, and
• customizing the structural components of the graph.

How to set default styles of Custom Graph Objects:
    By selecting Object Styles from the graph pull-down menu Insert, you may specify the default styles of custom text, shapes, arrows and pictures. The respective dialogs can also be accessed by highlighting a custom graph object and pressing ALT+ENTER (or selecting Object Properties from the Edit pull-down or flying (right-mouse-button) menus). When the Set as default style box is checked, the settings will become the defaults for all new objects of that type.
    Note that the Set as default style option is dimmed and cannot be de-selected when this dialog is called when an object is not selected, since then the only purpose for accessing this dialog is to specify new default settings. Therefore, any changes to the settings in this dialog will become the default settings.

How to add a new Plot to an existing Graph:
    Use the New Plot option in the graph pull-down menu Layouts (or the flying menu accessible by clicking the right-mouse-button on the Graph Data Editor). A subsequent dialog will allow you to specify the plot to be added. A new column (or a double-, or triple-column, depending on the plot type that you requested) will be added to the Graph Data Editor making room for data to be represented in that new plot. You can enter the data, paste them, or merge them from another graph.

What is the General Layout dialog?
    The General Layout dialog is one of two main graph customization dialogs (the other is the Plot Layout dialog, see the next topic).
    The General Layout dialog contains options to control and customize all those features and attributes which apply to the entire graph and are common to all component plots in the graph (see the section on differences between graphs and plots). They include such features as titles, gridlines, global colors (backgrounds, etc.), scaling, axis labels, or categorization labels.
    How to access the General Layout dialog for the current graph. The General Layout dialog is accessible from the pull-down menu Layouts, as well as via flying menus (which can be called by clicking the right-mouse-button on the graph; it is the first option in the "main" flying menu, accessible when you click the right-mouse-button on the empty area outside the graph axes). However, the quickest way to access it is to double-click (left-mouse-button) on the empty area outside the graph axes.
    Mixed-type graphs. As mentioned before, a graph may consist of plots of different types (see graphs vs. plots). Note, however, that one of the controls in the 2D General Layout dialog allows you to change the type of graph (see the list box Graph Type with graph-type icons). This graph type control is set to a specific type only if all plots comprising the graph are of the same type. Otherwise, it is set to Mixed, which means that different plots are of different types. By changing the Mixed setting in this control to a single specific type, all component plots will be changed to this particular type.
    Different formats of the Plot Layout dialogs. The format of the Plot Layout dialogs is somewhat different for each of the five main categories of graphs which can be produced in STATISTICA (see the next topic).

What is the Plot Layout dialog?
    The Plot Layout dialog is one of two main graph customization dialogs (the other is the General Layout dialog, see the previous topic). It contains options to control and customize all those features and attributes which apply only to specific plots and not necessarily to the entire graph (see the section on differences between graphs and plots). They include such features as type of plot, text of the legend, patterns, colors, and sizes of the plot representation, fitted function, confidence intervals or area settings for this plot, etc.
    How to access the Plot Layout dialogs for specific plots. The Plot Layout dialogs (one for each of the plots in the current graph) are accessible:

• from a list of plots in the pull-down menu Layouts;
• by double-clicking on the respective plot column header (or using the flying menu), or the Plot Layout toolbar button in the Graph Data Editor; and
• directly from the graph via flying menus (if you click the right-mouse-button on any part of a specific plot, the flying menu will list its respective Plot Layout dialog as one of the options).

How to adjust the margins of a graph:
    Margins within the graph area. Pressing the graph area and margins button on the graphics toolbar allows you to adjust the space between the edge of the plotting area (i.e., the borders of the graph window) and any graph components or custom graphic objects.
    Printout margins. The printout margins (the width of the distance between the edge of the paper and the beginning of the graph area) can be adjusted in the Print Preview dialog (pull-down menu File).

How to change the proportions (aspect ratio) of the graph window:
    The Graph Mapping Options dialog (accessible from the pull-down menu View) allows you to adjust the default graph window aspect ratio (e.g., you can set it to 1:1), and other general features of the graph area (margins, etc.); default settings can also be adjusted in that dialog.
    Also, interactive facilities can be used. Use the respective graphics toolbar buttons. Normally, the first of these buttons (MAR mode -- Maintain Aspect Ratio) is depressed, thus the proportions are automatically maintained as you re-size the graph. To change the proportions, press the second of these buttons which enables the free (non-proportional) graph window re-sizing mode. In this mode, the graph can be "stretched" or "squeezed" in one direction only, which changes the proportions between the X and Y coordinates of 2D displays and other relations between graph components. The modified proportions of the graph will be reflected in the printout (as can be examined using the Print Preview facility by pressing the Print Preview button or selecting the option in the pull-down menu File). After achieving the desired proportions of the graph, it is advantageous to switch back to the MAR re-sizing mode (press the first of the three buttons), so that the proportions are not inadvertently altered when re-sizing the graph.
    Press the third of the three buttons to enable the graph area and margins adjustment mode. In this mode, like in the non-proportional graph window re-sizing mode (see above), the proportions of the entire graph area can now also be altered, however, it will not change the size or the aspect ratio (X:Y) of the graph itself but only add or cut some space of the surrounding graph area.

How does the "Snap to Grid" facility work?
    The "Snap to Grid" facility, enabled by clicking the Snap to Grid toolbar button, aligns drawing objects with the drawing grid, an invisible network of lines that covers the drawing area (the grid may be displayed by pressing CTRL+G or selecting the option Alignment Grid from the View pull-down menu). When this facility is enabled, inserted objects are always pulled into alignment with intersections of alignment gridlines (the "snap to grid" effect). The default spacing of gridlines is 5% of the display; the vertical and horizontal spacing between gridlines can be changed.
    Initially, the grid begins at the 0,0 point in the upper-left corner of the graph window, but you can change the location from which gridlines are measured by editing the Horizontal and Vertical Origin values.
    When placing objects, you may toggle the snap to grid effect on and off by holding down the Tab key. In addition, you may specify whether the snap to grid alignment will apply only to objects placed in Fixed mode, or to all objects, regardless of drawing mode.

What are the Display Filters?
    This facility, accessible from the graph pull-down menu View, allows you to control the degree of "display filtering," that is, removal of overlapping points or text in the current graph.
    STATISTICA examines each point or text and does not draw the overlapping points or text within a designated radius (determined by the level of filtering) in order to increase the readability of the graph and speed up the graph redrawing process.
    The higher the level setting, the larger the filtering radius. In order to disable filtering, set the Display Filters to Off.

Dense scale values, point markers, or value labels are not displayed in the graph. How do I adjust the display to see them all?
    The display of overlapping items is suppressed by the Display Filters facility designed to increase the readability of the graph and speed up the graph redrawing process. In order to see all (overlapping) items, turn off the Display Filters (see the previous topic).

How to convert a graph into a data file:
    The simplest way to convert a graph into a data file is to highlight all data in the Graph Data Editor by clicking in the upper left corner, copy them to the Clipboard, (CTRL+C), then paste them (CTRL+V) into a data file.
    Also, STATISTICA graphics files (file name extension *.stg) can by imported as a "data" file using one of the data file import facilities provided in the Data Management module. The latter method will allow you to also transfer to the data file supplementary information, such as plot names and the graph title.

How to produce sequences of graphs from lists of variables:
    Specifying lists of variables for Stats Graphs. Most of the graph definition dialogs in Stats Graphs (accessible from the pull-down menu Graphs) allow you to select lists of variables in instances where a single variable is sufficient to define a graph. When such a list of variables is specified, STATISTICA will cycle through the list and produce one graph for each variable (e.g., a histogram or a line plot).
    Pressing the Next button (in the upper-left corner of the graph) will continue displaying the sequence of graphs; pressing the Quit button will interrupt displaying the sequence of graphs.
    "Cascades" of graphs requested from output dialogs. Most of the output (Results) dialogs in those statistical procedures which process lists of variables allow you to generate "cascades" of graphs for each (or each combination) of the variables in the current list. For example, such graphs can be produced from descriptive statistics, correlations, frequencies, crosstabulations, breakdowns, and other procedures:
    Pressing the Continue button (in the upper-left corner of the graph shown) will continue displaying the sequences of graphs. Pressing the "floating" Cont (continue) button with the name of the dialog from which the sequence of graphs was called (see the lower-left corner of the display shown above) will interrupt displaying the sequence.
    Automatic printouts. Note that when using either of the two methods mentioned above, one can automatically produce printouts of all graphs displayed on the screen by enabling the options Automatically Print All Graphs or Automatically Print all Graphs to Output Window (to direct them to the Text/output Window) in the Page/Output Setup dialog. When, instead of reviewing the graphs on-screen, you need to quickly produce hard copies (e.g., printouts), then it is advantageous to select the option Auto-Exit from Scrollsheets and Graphs in the Page/Output Setup dialog. If that option is selected, STATISTICA will "internally" press the Continue (or Next) button on every graph and Scrollsheet, thus allowing you to print long sequences of graphs (and Scrollsheets) without having to press a button at the end of every "queue-full" of output windows.
    Specifying sequences of graphs in SCL (STATISTICA Command Language) and STATISTICA BASIC (STATISTICA programming language). Finally, multiple graphs can be generated in batch using SCL (the command language) and STATISTICA BASIC (the data transformation and graphics language. SCL is well-suited to "programming" sequences of specific tasks including graphs which can be automatically generated in long series. STATISTICA BASIC allows you to custom-define graphs from any combination of data points in the data file (cases, variables); e.g., for exploratory data analysis, you can create one graph for every case of the data file. Sequences of those graphs (generated either by SCL or STATISTICA BASIC can be directed to the printer, the Text/output Window, or they can be saved to individual disk files.
    Note that sequences of graphs specified in either SCL or STATISTICA BASIC can be assigned to Auto Task Buttons for quick access at any point in your data analysis.

I have a free-hand drawn object on a graph and I would like to fill the patterns and make the object transparent. How do I do this?
    Select the free-hand drawn object and click the right mouse button. Choose the Fill Pattern option from the flying menu and select the On radio button under Display. In the Pattern combo box, select one of the H patterns (this refers to patterns that can be displayed as transparent patterns) and then select the Transparent radio button. Click OK.

How do I connect the midpoints in my Range Plot?
    Right-mouse click on any of the midpoints in your graph and select Change this Plot/series Layout. Change the Graph Type from Scatterplot to Line Plot and click OK. If you wish to display the point markers, press the Points button (in the same dialog) and set the Display to On.

When I create a histogram, the number of bins (bars) is different than what I would like to see. How do I change it?
    If you make this histogram from the Stats 2D Graphs—Histograms option available from the Graphs pull-down menu, then in the bottom right corner of the dialog, you can choose from several categorization methods. The default selection is probably not appropriate if your graph doesn’t look right. Try changing to one of the other methods, such as # of Categories.

Graph Customization - Scales

How to customize the layout and format of an axis (min-max, scale-values, tickmarks, etc.):
    Double-click on the respective axis to access the Scale Options dialog containing customization facilities for all features of the current axis.
    Alternatively, you can adjust individual features of the axis by selecting them from the flying menu accessible by clicking the right-mouse-button anywhere on the respective axis.
    Note that the applicable features of the axis can be copied to other axes by setting the respective switches in the dialog (see above; they can also be copied to other graphs, see the next topic).
    You can copy the features either to the corresponding (i.e., the opposite) axis or all other axes. The main scaling features of each axis can also be adjusted in the General Layout dialog, from which you can also call the Scale Options dialogs for each scale.

Can I copy or save axis definitions?
    Axis definitions with many settings and parameters (which are time-consuming to re-enter) can be re-used by transferring them using either the Clipboard or disk files to store axis definitions. Axis definitions for all axes or only the current axis can be saved or recalled from disk (file name extension *.axd), copied to or pasted from the Clipboard (using the options in the Transfer group in the Scale Options dialog).

How to replace numeric values on a scale with text value-labels:
    In the Scale Options dialog (see the previous topic), set the Scale Values to Text Labels. Then, in the editable list box Text Labels, enter the appropriate numeric values (determining where the text labels are to be placed on the axis) and the corresponding text-value labels.

How to insert a scale (axis) break symbol:
    You can place a "break" in a graph axis in order to show that the starting point (value) of the scale is arbitrary (e.g., it is adjusted to the observed minimum of the plotted values):
    To do this, double-click on the scale values of the axis in which you want the break to appear. In the resulting Scale Options dialog, select the option Place a 'scale break' on the axis and (optionally) adjust the break location. STATISTICA will place the break in the specified location of the scale after you click OK in the Scale Options dialog.

Can I shift the position of axes against the graph?
    In the Scale Options dialog, the Shift Axis option allows the position of the current axis to be shifted by a specified number of points (enter a positive value to shift the axis outside the graph, enter a negative value to shift it inside). For example, when X is selected as the current Scale, you can shift the X-axis upward by specifying a negative number as the Shift value.
    In order to show that there is no causal relation between the X and Y variables, shift both variables by specifying a positive value in the Shift field of the respective X- and Y-axes (causing a break where the X- and Y-axes meet). Note that this convention is not "universally" accepted, but is still used in some disciplines.

How to adjust the number of minor tickmarks (tickmarks between the scale values):
    STATISTICA will adjust the number of minor tickmarks to the current step size value. However, you can adjust the number of minor tickmarks (as well as the default style and size of minor tickmarks) for each of the axes with the Minor Tickmarks options in the Scale Options dialog. Double-click on the scale values of the axis in which you want the minor tickmarks to be displayed and select the desired number (from 1 to 15) of minor tickmarks. Selecting Auto will cause STATISTICA to select the optimum number of minor tickmarks. The default color and thickness of the tickmarks are determined by the current specifications of the respective scale line (see the Scale Options dialog).

What is the difference between the Manual and Auto (or Manual/0 and Auto/0) scaling of a graph axis?
    When the axis scaling is set to Manual, then the minimum, maximum, and step size for the axis are determined by the current values of the minimum, maximum, and step size (as entered in the respective fields of the Scale Options dialog. If it is set to Auto (i.e., automatic), then the program will automatically determine the scaling based on the range of values to be plotted.
    The Manual/0 and Auto/0 settings work like Manual and Auto (respectively), except that the steps on the scale will be determined as if they always started from 0.0 and not the minimum of the scale (regardless of whether the value of 0.0 is actually included in the current range of the scale or not). The scaling mode for each axis can also be specified in the General Layout dialog.

How to change case labels from numeric to dates on which the measurements were taken in a multiple line plot (these dates currently reside in variable 1)?
    Click the Options button in the 2D Line Plots Dialog and choose Var: Var 1 under the Display - Case Labels section.

I am creating a histogram of a discrete age-group variable for a subset of the cases in one of my data files. For the subset, the frequency for the lowest age group is zero. As a result, the STATISTICA histogram does not display that group on the x-axis. How can I force STATISTICA to display the empty age group?
    Double-click on the x-axis and the Scale Options: X dialog will open. Under SCALING, decrease the Min value (lower than the numeric value associated with the lowest age group).
    For example, if

How can I change the position of the scale values on the axes to be horizontal or vertical?
    Double-click on the axis label and the Scale Options: X dialog will appear. In the Value layout box under the Scaling section, make sure Normal is selected (if you want them to be vertical labels). Parallel will give you horizontal labels. When this is set to Auto (the default), the program will choose the appropriate format depending on the number of labels and space available.

Graph Customization - Titles, Legends, Custom Text

How to customize the location and format of the legend:
    Fixed vs. movable legends. Legends can be treated in two ways in STATISTICA graphs: either as fixed (unmovable) legends or movable legends. By default, when a graph is created, the legend is fixed (unmovable), which means that its position is automatically determined and the graph is moved to the left in the window to leave space for the legend (see below). You can make the graph legend movable so that you can reposition it in the graph and customize the text (e.g., add text beyond the 20 character limit, adjust line spacing and the distance between the legend symbols and the text, etc.) and other attributes of the graph legend by selecting the Move Legend option from the right-mouse-button flying menu (click on the fixed legend with the right-mouse-button, see below). When you select this option, the legend text will become like any other added text in the graph and you can edit the text or reposition the movable legend in the graph (click on it once and then drag it to the new position on the graph).
    Fixed (unmovable) legends can be temporarily removed from the graph by selecting the Fixed Graph Legend OFF option from the right-mouse-button flying menu. The fixed legend(s) can then be placed back into the graph (in its default position) by selecting the Fixed Graph Legend ON option available from the "main" flying menu accessible by clicking the right-mouse-button anywhere in the empty space outside the graph region.
    Superscripts, subscripts, and all other text format adjustments can be added to fixed legends using Control Characters in the appropriate Plot Layout dialog, but (unlike the movable legend) fixed legends are limited in the number of characters that can be displayed in the legend (20).
    The fixed legend font can be changed globally by double-clicking on the fixed legend or by selecting the Change Font option in the right-mouse-button flying menu.
    Moving and reformatting the legend. By default, the legends are automatically formatted and placed in one of the fixed (typically most space-saving) locations of the graph area.
    For most graphs, the default legends are formatted as columns of items positioned on the right side of the graph; each legend symbol can be followed by up to two lines of text with up to 20 characters each, and the texts can be edited in the Plot Layout dialog for the respective plot.
    The Move Legend option to customize the location of the legend (accessible from the legend flying menu) will turn off that default legend mechanism and convert the legend into a block of custom text. This text will have the same status and properties as text pasted onto the graph or entered using the custom text tool. Thus, you can move the legend to any position on the graph and take advantage of all custom-text formatting facilities.
    If you wish, you can include both fixed (unmovable) and movable legends in the same graph (see below). First, make the fixed legend movable by selecting the Move Legend option. Reposition this legend and then click on the graph background with the right-mouse-button and select the Fixed Graph Legend ON option from the flying menu. After you select this option, you will be asked if you want to remove (delete) the movable legend before placing the fixed legend in the graph. If you answer No, then the movable legend will remain and the fixed legend will be placed in its default position in the graph. If you answer Yes, then the movable legend will be deleted from the graph before the fixed legend is replaced.
    You can add a custom background, frame, place all items in one long line (by deleting the line feed characters from the end of each original line), or split the text into separate parts (select and cut a part of the legend, then paste it somewhere else).

What other types of legends are automatically created in graphs?
    In addition to the standard fixed legend (which identifies patterns and colors used to mark individual plots in the graph), there are also other more specialized types of fixed legends. For example there are contour legends which identify the levels in surface or contour plots, icon legends which identify the assignment of icon features to specific variables, or selection legends which identify the case selection conditions used to classify cases into multiple subsets shown on the graph. All these fixed legends can be changed to movable legends following the same conventions outlined in the previous topic. Also, the Fixed legend ON option from the graph right-mouse-button flying menu will enable all fixed legends which were turned off.

How to add a title to a graph:
    In every graph, there can be up to 11 lines of standard title text (5 on the top, 2 on the bottom, and 2 on each side), which are automatically positioned and centered.
    They can be edited in a dialog accessible by double-clicking on any title (you can access any one of the 11 lines of title in that dialog). The titles can also be edited in the General Layout dialog (see also the next topic).
    Note that all text entry fields in STATISTICA graphs support the formatting Control Characters (see below), thus titles can contain different text attributes (e.g., superscripts, italics), legend symbols, or text of any of the currently fitted equations.
    The effective number of lines for each of the types of titles can be increased by entering the line break Control Character (@A) or pressing the New Line button on the mini-toolbar.
    All other standard Control Characters are supported. For example, if you enter @F[1] in the title, then STATISTICA will replace this control character with the text of the current equation fitted to plot number 1 in the graph.

Can I convert the standard titles into movable text?
    Yes, by double-clicking on the title to edit it, then cutting the text to the Clipboard (by pressing CTRL+X or the Cut toolbar button), then pasting the text into the graph (press CTRL+V or the Paste toolbar button). The text will appear in the default, upper-left corner of the graph, from which it can be moved by dragging with the mouse.

How to place a graph title or a footnote in a fixed position:
    You can convert the standard title into moveable text (see above) and then fix it in the desired location. Also, you can create new text, as explained below. Use the Graphic Text Editor to enter and position the title or footnote in the desired location.
    If you intend for the text to stay in a particular place in the graph area regardless of the future changes to the graph scales or graph location (within the graph area), change the status of the object from the default Dynamic to Fixed, which will keep the text in the absolute window coordinates regardless of the changes to the graph (e.g., in 5% of the window width and length from the upper left corner). See also the previous topic.

What are Control Characters?
    Control Characters are specific characters that can be entered in graphic text (e.g., titles, legends, scale values, added text, category values, category names) in order to customize the text.
    If the text is entered via the Graphic Text Editor dialog, then specific customizations (e.g., attributes of text) can be entered by highlighting the respective part of the text and then pressing one of the mini-toolbar buttons. When you click on these buttons, STATISTICA will automatically insert the respective Control Characters into the text.
    Both in the Graphic Text Editor window and in all other instances where the graphic text can be entered (e.g., in titles, legends, etc.), the Control Characters can by typed directly into the text. Some of the most commonly used Control Characters are listed below:

How to change the font for a part of the custom text:
    Select the text, then press the Local Font button, and in the Font dialog select the desired font. The Control Characters @K and @Z will be added before and after the selected text.
    Technical Note: Representation of local (user-specified) font. When a custom font is selected for part of the text, an additional font description is added to the internal structure of the custom text object and assigned a sequential number. The Control Characters surrounding the custom text refer to this internal "font list;" for example, @K[1] (or @K) would refer to the first additional font in the list; @K[2] would refer to the second additional font. A previously specified font selection in the same custom text object may be repeated by referencing the same control character (e.g., @K[3]...@Z if it was the third, custom-defined, local font). To determine the font style set for a portion of the text, position the cursor in that part of the text and press the Local Font button.
    Inserting symbols. In order to select a special symbol or character, press the Symbols button, which will bring up the Windows Character Map facility (by default, the Symbol font is selected).
    You can use the Windows Character Map to insert into documents special symbols or characters not found on most keyboards. These include special characters provided in symbol fonts. Double-click the desired character (or press the Select button) to move the respective character to the edit field (called Characters to Copy) of the Character Map facility. The character(s) selected in the Windows Character Map are copied to the Clipboard when you press the Copy button in the dialog; they may then be pasted into STATISTICA (or another application). Note that the Windows Character Map transfers only the ASCII representation of the characters to the Clipboard; you need to select the respective font in your application in order to properly display the character, that is, highlight the pasted characters and press the Local Font button to select the proper font. For more information on the Windows Character Map, press Help while you are in that utility.

How to format text (e.g., use italics, sub- and superscripts, etc.) in graphs:
    If the text is entered into the Graphic Text Editor, highlight the part of the text to be changed and press the respective mini-toolbar button (e.g., bold). If the text is entered outside the Graphic Text Editor, then use the Control Characters). The same Control Characters (i.e., @I, @D, @U, etc.) can be used to format and customize all types of text in STATISTICA graphs (e.g., titles, value labels, category names and values, scale labels, custom text, etc.).

How to rotate text:
    You may select the orientation (Horizontal, Vertical Left, or Vertical Right) of text objects (custom text and moveable legends) in the graph in the Orientation section of the Graphic Text Editor. You can also choose to rotate the text by specifying the rotation angle (from 0 to 359° or 0 to 359°) in the Angle edit field.
    Alternatively, you can interactively rotate the text by selecting it in the graph and then dragging one of the object handles (small black squares) in the desired direction, or by using the hot keys PAGE DOWN and PAGE UP to rotate text objects selected in the graph clockwise or counterclockwise, respectively, in 5° increments. To rotate in 1° increments, hold down the CTRL key while pressing PAGE DOWN or PAGE UP.
    The rotation of text objects takes place around the object's anchor point, which is visible as a "cross-hair" when the object is selected. The position of the anchor point may be adjusted in the Graphic Text Editor. The rotation angle (in degrees) will be displayed in the Change Show window of the graphics toolbar.
    Technical Note: Rotatability of fonts. Some fonts may not support rotation, and some fonts may support only limited rotation and will approximate rotation to the nearest angle to which it may be rotated. When these fonts are used, misalignment between the frame around the text and the font may occur. In this case, do not place a frame around the text. Also note that some printer drivers may not support rotation of some fonts (even though the text may appear properly rotated on screen).

Fitting, Plotting Functions

How to fit a function (line or surface) to data:
    Access the Plot Layout dialog for the respective plot; then, select the desired type of function or smoothing procedure to be used from the list box FIT (note that for some 3D graphs the Plot Layout and Graph Layout dialogs are combined). You can adjust the fitting options (e.g., stiffness or optimization settings) and the pattern for the graphical representation of the fit by pressing the Options and/or Pattern buttons, respectively (they are located in the FIT area of the dialog). The pattern can also be adjusted by double-clicking on the fit line or surface in the graph. (See also the section How to Fit a Custom-Defined Function to the Data.)

How to display a specific equation for the fitted function (e.g., a polynomial function):
    You can use the Control Characters for the fitted function in a specific title or in the Graphic Text Editor in order to place the fitted function in the graph. The Control Characters are @F[plot number, subgraph number] where you specify the desired plot number and optionally the subgraph number (for categorized graphs). For example, if you enter the Control Characters @F[2] into the fixed title of the graph, then STATISTICA will enter in that title position, the fitted function for the second plot in a multiple plot graph (i.e., a 2D line plot with three plots). If you enter the Control Characters @F[1,3] in a title, then STATISTICA will place in that title position the first fitted function for the third subgraph in a categorized graph. For other options and details, see the Electronic Manual.
    Displaying the text of fitted function equations in Stats Graphs. In Stats Graphs (available from the pull-down menu Graphs), the display of the text of the fitted function equations can be requested by selecting the Display Fitted Function (equations) option in the Options dialog. (That dialog controls settings which are common for many types of Stats Graphs and it is accessible by pressing the Options button in every Stats Graphs definition dialog.)
    In all one-plot and non-categorized graphs where only one function is fitted, the text of the equation is displayed in the first available line of the fixed title. Depending on the number of equations to be displayed, also in categorized graphs, the equations can be displayed in the fixed titles of the graph (typically four lines are available if the custom Job Title or Graph Title options are not used, see the Options dialog).
    However, if more equations need to be displayed than the number of lines available in the fixed title, then STATISTICA will create a custom text object on the graph and place the equations there. Potentially, such lists of equations may be very long (e.g., include 256 equations), and thus the custom text object may be large and partially cover the graph. However, the location of the listing of functions can be adjusted (the list can be moved around and edited like any other custom text object, the font size reduced, etc.).
    When the listing of functions is very long, it is recommended to add some space around the graph and place the text object there.

How to plot a custom-defined function:
    Select the 2D or 3D Custom Function Plot option in the pull-down menu Graphs and specify the function in the respective dialog. Also you can add a custom function plot to any existing graph, as explained below.
    First, access the Plot Layout dialog for the respective plot (e.g., use the graph pull-down menu Layouts or a flying menu; note that for some 3D graphs the Plot Layout and Graph Layout dialogs are combined). In the Plot Layout dialog, press the Custom Function button to access the custom-equation editor and specify the equation to be plotted in the 2D or 3D graph.
    In addition to the standard math functions, a variety of functions representing distributions as well as their integrals and inverses are supported and can be plotted (including beta, binomial, Cauchy, Chi-square, exponential, F, gamma, geometric, Laplace, logistic, normal, log-normal, Pareto, Poisson, Student's t, and Weibull distributions).
    Press the Help button or the F1 key to access the detailed syntax description and examples in the Electronic Manual.

How to fit a custom-defined function to data:
    The custom-function plotting facility (see the previous topic) accessible in the Plot Layout dialog will only plot the requested (custom-defined) functions and overlay them on the existing graph, but it will not fit these functions to the data. A selection of the most-commonly used, predefined functions which can be fitted to the data and smoothing procedures is available from the same dialog (e.g., linear, logarithmic, exponential, various polynomial, distance-weighted least squares, spline, and others); see the section How to fit a function (line or surface) to data.)
    Comprehensive facilities to fit to data (and interactively plot in two or three dimensions) user-defined functions of practically unlimited complexity are provided in the Nonlinear Estimation module.
    For more information on those techniques, refer to the Electronic Manual.

How to change the axis-proportions (aspect ratio) in 3D graphs:
    The Graph Mapping Options dialog (accessible from the pull-down menu View) allows you to adjust the default graph aspect ratio (e.g., you can set it to 1:1), and other general features of the graph area (margins, etc.).
    Note, however, that unlike 2D graphs, where changes to the aspect ratio require adjusting a relation between only two measures (X:Y, and thus can be accomplished by non-proportional graph window re-sizing), the aspect ratio for 3D displays is defined by three parameters (X:Y:Z).
    The non-proportional window re-sizing mode will not affect the axis proportions in 3D graphs, it will only result in adjustments of the margins of the plot area. In order to adjust the axis aspect ratio, press the More Options... button in the General Layout dialog. Use the Axis Proportions section of the 3D Graphs/More Options dialog to make the adjustments.
    Note that the adjustment does not modify the proportions of the graph window (only the axis proportions of the graph are modified).

Brushing, Interactive Analysis

Glossary of Brushing terms:
    The following terms are used to denote specific aspects of brushing operations.

• Highlighted: The initial state of selected data points prior to performing some brushing Action (e.g., labeling, see the term Action, below). Highlighting data points in the graph works like in word processors or spreadsheets where highlighting is a precursor to performing an Action on the highlighted text (e.g., applying a font attribute such as boldface or italic). If no Action is taken and you exit the brushing mode, the highlighted status of a point is hidden (not displayed), but preserved internally (e.g., is saved with the graph) and the points will be highlighted once again whenever you re-enter the brushing mode for the graph.
• Action: An operation (marking, labeling, or turning off) performed on highlighted points. Actions are executed when you press the Update button in the Brushing dialog (or after every operation of pointing to a data point with the mouse when you are in the Auto Update mode).
• Selected: A state in which points have been Highlighted or further identified by execution of some Action. The selected status of a point is saved with the graph. Four selection attributes are available:
  • Marked: Identification of selected points by displaying an alternative point symbol.
  • Labeled: Identification of selected points by displaying their respective data point labels.
  • Turned off: Temporary removal of points from the display and fitted function calculations. The point values remain visible in the Graph Data Editor.
  • Highlighted: Identification of brushed points which is available only when viewing the graph in brushing mode. This status is usually a precursor to some further Action, such as Labeling, Marking, or Turning off.
• De-selecting: Removal of all attributes (marked, labeled, turned off, and highlighted); this is the default status of all points when a graph is created.

How can brushing be used in exploratory data analysis?
    There are countless applications of brushing to explore relationships between variables, and/or the contribution of specific data points or subsets to those relationships. A "typical" illustration of the use of brushing in exploratory data analysis is the examination of the contribution of data points representing different ranges or different levels of one variable to correlations between other variables, which can be visually inspected using a scatterplot matrix.
    For example, by including a categorical, three-level variable such as income level (Incomlvl) in a scatterplot matrix (see above), then by using the rectangular or lasso brush, points can be selected from one income level, and the location of these points in scatterplots of all other variables (e.g., Assets, Debt) becomes immediately apparent.

What is Animated Brushing?
    Typical applications of Animated Brushing are in exploratory data analysis using matrix plots (see the previous topic), where instead of brushing consecutive ranges of a variable (to explore the influence of various sections of its distribution), you can invoke an automatic movement of the brush (a rectangle or lasso brush) and watch the "results."
    Specifically, a brushing region is defined in one subgraph in the matrix and is automatically moved across the subgraph (horizontally, vertically, or both). As the brushing region passes over groups of points in the subplot, corresponding points in all other plots are highlighted.
    The Animation option is available from the Brushing dialog whenever the rectangle or lasso brush option is selected.

How to interactively review points which belong to specific plots in multiple scatterplots and other multiple graphs:
    Identifying all points of a plot. In the (default) pointing mode (when the Point Tool is enabled), click with the left-mouse-button on any point that belongs to the specific plot and all points of that plot will become highlighted. They will stay highlighted for as long as you keep the mouse-button pressed. If there are many plots in the graph, and their respective point markers are small and difficult to identify, then you can click on the legend (with the left-mouse-button). This will also highlight all points that belong to the respective plot.
    Identifying individual points of a plot. If you need to identify values of specific points in the graph, use the Brushing Tool.

How to identify (in graphs) specific subsets of data:
    Stats Graphs (accessible in the pull-down menu Graphs) offer facilities to define subsets of cases to be identified in graphs. User-defined "multiple-subset" definitions of such subsets can be entered as logical case selection conditions of virtually unlimited complexity using facilities identical to those illustrated in producing categorized graphs.
    These subset identification facilities are supported in many types of Stats Graphs, including matrix plots, icon plots, 2D scatterplots, 3D scatterplots, 3D trace plots and other graphs.
    Subsets may also be identified by using the Brushing - Extended Options (available by clicking the More button in the Brushing dialog). This facility offers a wide variety of methods to identify subsets of data. In this dialog you may select ranges of values for the variables in the plot, and examine their relations to other variables (for example, in a matrix plot).

Can I interactively identify data points in a graph?
    Yes, by marking them in the Graph Data Editor, or labeling them selectively (via Brushing).

3D Displays - Interactive Analysis

How to rotate a 3D graph:
    Either use the Rotation and Perspective control button on the graphics toolbar to access the interactive rotation control facility, or press the More Options... button in the General Layout dialog, to enter specific viewpoint parameters controlling the position of the imaginary viewpoint against the 3D object.

How to adjust the perspective ("viewpoint") in a 3D graph:
    Either use the Rotation and Perspective control button on the graphics toolbar to access the interactive perspective adjustment facility, or press the More Options... button in the General Layout dialog, to enter specific viewpoint parameters controlling the position of the imaginary viewpoint against the 3D object.

How to perform exploratory spinning of 3D datasets:
    Clicking on the Rotation and Perspective toolbar button brings up the Perspective and Rotation window allowing for rotation, spinning (for analytic or exploratory purposes) and interactive adjustment of the point of view for three-dimensional displays.
    By clicking on the Spin button, the display can be set into continuous rotation in clockwise or counter-clockwise directions. Note that a simplified representation of the graph data is displayed initially (to reduce the redrawing time). All data points will be shown, however, if you request the Spin function, so that the complete pattern of data can be explored.
    Refer to the Electronic Manual for more information about perspective and rotation controls in this window. Note that specific adjustments of the viewpoint and perspective (e.g., for an exact reproduction of a display) can also be made by editing the numeric viewpoint parameters accessible by pressing the More Options... button in the General Layout dialog (use the pull-down menu Layouts or double-click on an empty area of the graph outside the axes).

How to review cross-sections of a 3D graph:
    An Animated stratification option is accessible by pressing the respective button on the toolbars of all 3D sequential graphs. A dialog will appear on the top of the graph allowing you to control the display of consecutive "slices" of the display.

How to use the "X-ray" facility to explore layers of 3D graphs:
    In addition to the animated stratification option (see the previous topic), in the default pointing mode (when the Point Tool is enabled), you can selectively highlight individual plots in the graph by pressing the left-mouse-button anywhere on the selected plot.
    For example, in sequential 3D plots, this option allows you to temporarily highlight complete series of data even if the respective plots are almost entirely covered by other plots (as if you were "X-raying" a plot).
    If the plot of interest is covered entirely, click on its legend to achieve the same result. This facility is also useful in examining other types of 2D and 3D graphs, as it allows you to uncover invisible areas of specific plots or discriminate between different plots (e.g., in multiple scatterplots). In 3D histograms, which belong to a "one-plot only" category of graphs (see plot vs. graph), the selection facility can also be useful to review the hidden layers of the graph because in those specific graphs (3D histograms), the tool is automatically set to highlight individual layers and not separate plots.

How to define a custom selection of levels for a contour plot or surface plot:
    Double-click on the surface of the graph to access the 3D Graphs/More Options dialog (also accessible by pressing the More Options... button in the General Layout dialog). For Surface/Contour Levels, select User-Defined and press the Specify Contours button to display the Specify Contour Levels dialog.
    From this dialog, custom surface/contour levels may be defined, and level definitions may be saved to the disk (to be used in other graphs). Facilities are provided to automatically generate sets of levels following the user's specifications (see Interval Levels, Uniform Levels). Default settings can be adjusted in the Global Defaults dialog accessible from the pull-down menu Options.

How to define a custom palette for a contour plot or surface plot:
    Double-click on the surface of the graph to access the 3D Graphs/More Options dialog (this dialog is also accessible by pressing the More Options... button in the General Layout dialog). Select User-Defined and press the Specify Contours button to display the Specify Contour Levels dialog (see previous topic). Press the Get Palette button to bring up the Edit Palette dialog.
    Now, you can build a custom palette by inserting, deleting, adding, and replacing colors. As you build the palette, you can press the Custom Colors button to define custom colors whenever necessary.
    The Edit Palette dialog can also be used to save and retrieve palettes to/from disk files (the default file name extension for palette files is *.pal).

Compound Graphs, Embedding, Multiple Graph Management, Clipboard

How to place one STATISTICA graph into another:
    Pasting one graph in another. The easiest way to place one graph into another is to copy a graph displayed in one window (press CTRL+C, press the toolbar button Copy, or use the option Copy graph to Clipboard in the "main" flying menu accessible by clicking the right-mouse-button anywhere on the outside of the graph axes). Then, move to the target graph window and paste it there (press CTRL+V, press the toolbar button Paste, or use the flying menu). The pasted graph will appear in the upper-left corner of the target graph. Now, you can move or re-size it like every other custom graphic object.
    You can also change the properties of the pasted object by selecting the Object Style option from its respective flying menu (click on an object with the right-mouse-button) or pressing ALT+ENTER (the Object Properties key). You can also edit the embedded object by double-clicking on it (following the standard OLE conventions).
    Linking and embedding. Graphs and artwork saved into disk files can also be dynamically linked or statically embedded in the current graph by using the standard OLE facility, accessible by pressing the graph toolbar Insert button.
    Technical Note: Mapping of inserted STATISTICA graphs. If the inserted object is a STATISTICA graph, several graph mapping options are available, and they can be adjusted in the Object Properties dialog (accessible from the Paste Special or Insert dialogs, before the graph is inserted, or later from the right-mouse-button flying menu for the object or by pressing the (Object Properties) ALT+ENTER key when the object is selected). For example, if you select the Fixed Size Font (Mapping mode) setting, then the physical size of the font in the object graph will be tied to the Mapping Base of the master window, and thus, it will be automatically adjusted when you press the Global Font Size buttons for the graph.

How to place a foreign graph or artwork in a STATISTICA graph:
    The Clipboard-based (including the linking and embedding, OLE) as well as inserting operations listed in the previous topic apply to all Windows-compatible graphs and artwork. Linking and embedding operations support graphs and artwork saved into bitmap-format files, Windows graphics metafiles, STATISTICA format graph files, and any OLE-compatible objects.

How to place text (e.g., reports, tables, etc.) in a STATISTICA graph:
    Even large portions of text (e.g., a report several pages long) can be pasted into STATISTICA graphs using the Clipboard operations mentioned in the previous two topics. The text can then be edited and customized (within STATISTICA graphs) using the STATISTICA Graphic Text Editor or the respective OLE-server application.
    Both the Clipboard-based as well as inserting operations listed in the previous topic apply to all Windows-compatible graphs and artwork (linking and embedding operations support any OLE-compliant objects).

What are compound graphs?
    Compound graphs are those that contain linked or embedded OLE objects. STATISTICA can automatically create compound graphs (e.g., in the Quality Control module where one display contains 4 different types of graphs, or when you use the Multiple-Graph AutoLayout Wizard). STATISTICA supports a comprehensive selection of facilities to insert OLE-compliant objects in STATISTICA graphics documents (for more information, see the next topic).
    One of the unique features of STATISTICA's OLE implementation is that it supports nested (up to fourth order) compound documents, that is, you can embed in a STATISTICA graph, a graph containing embedded objects, with embedded objects.
    Another unique feature of STATISTICA compound graphs is that if the component objects are its own (STATISTICA) graphs, the mapping of the graph objects can be adjusted to optimize the appearance of the fonts, markers, and spacing between parts of the graph in relation to the overall size and proportions of the master document (see Mapping of inserted STATISTICA graphs).

Representing objects in graphs as expandable icons:
    Icons representing documents in the File Manager can be dragged across applications and dropped into STATISTICA graphs. If the source application is OLE-compliant, the document will appear in the STATISTICA graph.
    If the source application is not OLE-compliant, then the document will be represented as an icon, either of the source application (if an association exists in Windows for the document's file extension), or of the Windows Object Packager (if no association exists).
    These icons function as buttons; double-clicking on an icon will launch the application with which it is associated and open the file represented by the icon.

How to copy an entire STATISTICA graph to the Clipboard:
    Make sure that the window containing the graph to be copied is active, then press CTRL+C. Alternatively, you can press the toolbar button Copy, or select the option Copy graph to Clipboard in the "main" flying menu accessible by clicking the right-mouse-button anywhere on the outside of the graph axes.
    STATISTICA graphs can be pasted and linked or embedded in other application documents (e.g., word processor documents, spreadsheets) following the standard OLE conventions. If STATISTICA graphs are pasted to OLE-compatible applications, the graphs maintain their relation to STATISTICA and thus can be interactively edited from within the other application, or updated when the STATISTICA graphs change.
    If the STATISTICA graph copied to the Clipboard had been saved (to a *.stg file), you can link it in other application (or STATISTICA's own) documents by using the Paste Link option from the pull-down menu Edit (or Paste Special).
    Technical Note: Resolution of copied graphs. Metafile representations of STATISTICA graphs (one of the formats created in the Clipboard and commonly used by other Windows applications) can be produced either at the resolution of the currently specified printer (e.g., 600 dpi), or at the resolution of the screen display. The resolution used is determined by the Metafile Mode option in the Page/Output Setup dialog.
    Usually, Printer resolution is desired for creating hard copy output, as this will produce the best results in the final copy. However, because the resolution of the printer generally exceeds the screen resolution, some dithering of the image will result in the display. In order to paste a graph in an another application only for screen display purposes, you should select Screen Resolution for the best results.

How to copy a selected part of a STATISTICA graph to the Clipboard:
    Copying an object. Select a graphic object to be copied by clicking on it (be sure that you are in the default pointing mode, i.e., the Point Tool button on the toolbar is pressed). Graphic objects are all objects you have created on the screen, such as a custom text, a segment of a drawing, or an embedded graph or artwork). When the object is selected (highlighted), press CTRL+C. Alternatively, you can press the toolbar Copy button.
    Copying a rectangular section of the graph. Enable the Screen Catcher tool by pressing ALT+F3 (or selecting it from the pull-down menu Edit or the "main" flying menu options accessible by clicking the right-mouse-button anywhere on the outside of the graph axes). The cursor will change to a small circle with a cross hair; place the cross hair in the upper left corner of the area to be copied, then drag it to the lower right corner (a rectangle will indicate the exact area which you are selecting). When you release the mouse button, the selected area will be automatically copied to the Clipboard in the bitmap format (there is no need to press the Copy button). Note that the Catcher can be used to copy any rectangular part of the screen, not only in the graph window from which it was called but any part of the screen (even including parts that belong to other applications).
    When selecting the area to be captured, you can rotate by 180°, and/or reverse (produce a mirror image of) the area captured to the Clipboard by choosing one of four directions of dragging the mouse. The area captured with the Screen Catcher facility will be rotated (which can be achieved by dragging the mouse in different directions over the area to be captured) and displayed in the Clipboard.

What is the Screen Catcher?
    The Screen Catcher is a STATISTICA utility which can be used to copy any part of the screen to the Clipboard (for details, see the second part of the previous topic).

How to create a "blank" graph (space for drawings or artwork to be pasted or linked/embedded):
    Select the Blank Graph option from the Graphs pull-down menu or the Graphs Gallery in order to open a blank graph. Here, you can add new or existing graph objects (e.g., added text, embedded or linked objects, arrows, free-hand drawings, previously saved graphs, etc.).
    The Multiple-Graph AutoLayout Wizard (see the next topic) and the Templates option (from the Graphs pull-down menu) can be used to design and produce a custom layout. Alternately, the Snap to Grid facility can be used.
    Also, the Alignment Grid (accessible from the Graphics window pull-down menu View) and/or the dynamically-updated cursor coordinates (in the Show Field segment of the graphics toolbar) can be used to aid in the visual placement and alignment of the graph objects in the blank graph.
    Blank graphs can be created in the default or printer page proportions. When you select the Default Proportion option, the proportions of the blank graph window will follow the default graph window properties (i.e., it will not necessarily be the same as the current printer/page proportions, e.g., portrait/landscape, margins, etc.). The graph proportion can later be adjusted via the Use Print Proportion option from the View pull-down menu (so that the graph will fill the physical size of the printed page) or the Change Proportions option in the View pull-down menu.
    Selecting the Printer Page Proportions option will open a blank graph window which is proportional to the current printer page proportions, as specified in the Print Preview, Margins, and Printer Setup dialogs (e.g., portrait or landscape). The blank graph proportion will be automatically adjusted so that the graph will fill the physical size of the printed page.

How to place multiple graphs on one page:
    Several graphs can be printed on one page by linking or embedding them within a blank graph. Although this can be done manually using cut-and-paste (and Snap to Grid), the easiest method is to use either the Multiple Graph Layouts/Templates (see pull-down menu Graphs or Graphs Gallery), or the Multiple-Graph AutoLayout Wizard (see the next topic), which automates placement of multiple graphs on one page. See also Mapping of inserted STATISTICA graphs.

What is the Multiple-Graph AutoLayout Wizard?
    The Multiple-Graph AutoLayout Wizard may be accessed from the Multiple Graph Layouts option either in the Graphs Gallery (press the Graphs Gallery toolbar button) or in the Graphs pull-down menu. The Multiple-Graph AutoLayout Wizard assists you in selecting and arranging graphs to be placed on the same page.
    Graphs may be selected from all currently open STATISTICA graph windows (in all currently open STATISTICA modules), or from graph files previously saved to disk; blank graphs (to be filled or paste-replaced later) can also be used.
    Graphs may be combined with titles, subtitles, and footnotes (independent of text contained in the individual graphs), and automatically arranged in various ways, depending upon the number of graphs to be combined.
    Often it is desirable to rescale the fonts, markers, and other components of the graphs before they are reduced and placed in larger displays. This can be automatically accomplished in STATISTICA by selecting the Fixed Size Font setting of the Mapping Base.
    When all component graphs are set to Fixed Size Font, you can adjust all fonts, markers, and spacing (defined in points) simultaneously by using the Global Font Increase/Decrease toolbar buttons.

How can I undo operations on objects?
    A multi-level undo option (available from the Edit pull-down menu) maintains separate queues of previously performed operations for separate objects. The undo operation works on the currently highlighted object; therefore, if you press CTRL+Z, STATISTICA will undo the last operation performed on the currently selected object and not merely the last operation on an object. Note that this option can also be used to undo operations on linked or embedded objects.

How do I change the order in which my graphs appear in the Multiple Graphs Layout?
    In the AutoLayout Wizard - Step 2 dialog you can select to move (reorder) graphs in the Graphs: section.

Printing Graphs

How to print graphs:
    The quickest method to print the current graph is to press the Print button on the graphics toolbar, which is a shortcut method to print the graph to the currently selected output channel (Printer and/or the Text/output Window) following the default graph printout settings (or settings as they were last modified in the Page/Output Setup dialog ).
    If you need to modify any graph printout settings, use the Page/Output Setup dialog or the Print option in the pull-down menu File. Unlike the Print button on the toolbar, using the menu option will not initiate the printing immediately, but will first display an intermediate dialog (Print Graph) allowing you to adjust various printout and printer settings (see the next topic). Use the Print Preview option (the Print Preview button or pull-down menu File) to see how the graph will appear on the page and to interactively adjust the margins with the mouse (see the next topic).
    Automatic printing. If you want STATISTICA to automatically print every graph which is displayed on the screen, select one of the Auto-report options in the Page/Output Setup (Graphs) dialog shown above (accessible by double-clicking on the Output field on the status bar at the bottom of the STATISTICA window).
    Each graph will be automatically printed and/or sent to the Text/output Window without displaying any intermediate dialogs or asking you for confirmation. When this automatic graph printing facility is used with analyses which generate large numbers of graphs (e.g., cascades of 3D histograms for every "slice" of a multi-way table), then it is advantageous to select the Auto-Exit from Scrollsheets and Graphs option in the dialog. If that option is selected, STATISTICA will "internally" press the Continue button on every graph and Scrollsheet, thus allowing you to print long sequences of graphs without waiting for you to press the Continue button at the end of every "queue-full" of graphs.
    Batch printing. There is also a batch printing facility which can be used to print previously-saved graphs. It is available by selecting the Print Files... option from the pull-down menu File.

How to modify the graph printout settings (margins, resolution, dithering):
    Print Preview. If you wish to see the graph as it will appear on the page, or modify the size of the print area and the margins of the printout, use the Print Preview facility accessible by pressing the Print Preview graphics toolbar button, or in the pull-down menu File.
    The Print Preview window is resizable, thus you can see details of the graph as they will appear on the printout. The margins can be adjusted (after pressing the Margins button) by dragging the margin lines or entering specific printout size measurements (in inches, centimeters, or percentages).
    Graph printout settings. If you need to modify any graph printout settings, open the Page/Output Setup dialog (accessible by double-clicking on the status bar at the bottom of the STATISTICA window). From this dialog, you may click the Margins/Setup button to adjust the page margins around the graph.
    Clicking on the Printer Setup button allows selection of the output device (from among all devices installed via the Windows Control Panel), and specification of the orientation, paper size, and source. Note that some printer drivers support some of STATISTICA's advanced printer control features (e.g., background shading or text rotation) only when they are set to a higher resolution (e.g., higher than 150 DPI) and/or when they are set to print fonts as graphics (see the next topic for more information).

Do all printer drivers support rotated fonts? (Some rotated fonts appear horizontal on the printout even though they are properly rotated on the screen and in the Print Preview mode.)
    Most properly configured printers supported by Windows can properly handle rotated fonts, however, some printer drivers support some of the advanced printer control features used by STATISTICA only when they are set to a higher resolution (e.g., higher than 150 DPI) and/or when they are set to print fonts as graphics. If you encounter problems (e.g., rotated text is printed as unrotated or "uncovered" text which was supposed to be covered), first try to enable the Print TrueType as Graphics option in the Setup/Options dialog of the Printer Setup... (pull-down menu File). If this does not help, set the printer to a higher resolution (e.g., change 150 DPI to 300 DPI).

Do all printers support the non-transparent overlaying of graphic objects?
    Most properly configured printers supported by Windows can properly handle printing of non-transparent overlays used in STATISTICA graphs; see the previous topic for advice on how to configure the printer driver.

How to control the minimum line thickness in graph printouts:
    This setting can be adjusted in the Page/Output Setup dialog, and the minimum line thickness setting entered there will override any smaller line thickness settings specified for individual graph components.

How to make the light graph background appear in black and white printouts:
    Graph backgrounds may be printed as shades of gray and the darkness of the background color specified in the graph will be represented by a darker shade of the printed gray pattern. This option can be set in the Page/Output Setup dialog.

How to use different fill patterns to print solid lines of different color:
    This setting can be adjusted in the Page/Output Setup dialog (see the first topic in this section). The option to represent solid lines of different color as different patterns may produce desirable results when the lines to be printed are very thick (e.g., 5 points or more). However, this option is not recommended to be used with thin lines because the differences in patterns representing different colors will not be salient enough to improve the readability of the differences between the lines, and at the same time the lines may appear jagged.

What is the Graph Mapping Base?
    Technically speaking, the Graph Mapping Base is a coefficient which controls how the logical sizes of fonts, markers, and related sizes (as set by the user in points, 1 point= 1/72 of an inch) translate into physical sizes which appear on the screen or printout (see the next topic). The Graph Mapping Base coefficient can be adjusted in the Graph Mapping Defaults dialog, accessible from the Global Defaults dialog, or interactively via the Global Font Increase/Decrease toolbar buttons.
    The meaning of the Graph Mapping Base value can be explained as follows. Graph objects (e.g., fonts, point markers, etc.) of a specified logical size will be printed at the same physical size if the diagonal of the output (in points) is equal to the absolute value of the current Graph Mapping Base (as set in the Graph Mapping Defaults dialog, accessible from the Global Defaults in the pull-down menu Options). If the diagonal of the output is larger or smaller than the absolute value of the Graph Mapping Base, then the physical sizes of graphic objects will be proportionally larger or smaller, respectively, than their logical sizes. See also the next two topics.

Are fonts set to specific sizes (in points) always printed having the requested physical sizes?
    No, it depends on the current setting of the Graph Mapping Base (see the previous topic), as well as the requested printout size.
    Dynamic image scaling in STATISTICA. In STATISTICA, all graph displays and printouts can be continuously scaled. STATISTICA will also automatically adjust the sizes of all fonts, markers, spacing, etc. (proportionately to the overall size of the graph), such that manual adjustments of individual font sizes are rarely necessary.
    The default printout size. Fonts will therefore be printed as having their specific physical sizes (as set in points; 1 point = 1/72 of an inch), only when the default setting of the Graph Mapping Base (see the previous topic) and the default printout size settings are used or when the length of the diagonal of the display or printout is equal to the current setting of the Graph Mapping Base. Otherwise, they can be printed smaller or larger depending on the requested printout size. Specifically, the fonts will appear printed in their exact physical size, i.e.:
    This is 8 point
    This is 10 point
    This is 12 point
    This is 14 point
    This is 16 point
    This is 18 point
    This is 20 point
if letter-size paper is used, in portrait orientation with default (1 inch) margins on all sides, and if the default graph proportions and Graph Mapping Base are not modified. Technically speaking, the Graph Mapping Base value is the length of the diagonal (in points) of the output at which fonts will appear at the same physical font size as their logical definition in the graph.
    The manner in which the font size settings as requested in the graph translate into the actual physical sizes of the fonts which are displayed or printed (i.e., the Graph Mapping Base) can be globally adjusted in the Graph Mapping Options dialog, accessible from the pull-down menu View. This can also be done interactively, using the Global Font Decrease or Global Font Increase buttons on the graphics toolbar (see the next topic).

How to quickly adjust (re-scale) sizes of all fonts in a graph:
    A global font size adjustment facility is provided which will proportionately adjust the sizes of all fonts in a graph (both on the screen and in graph printouts). Technically speaking, this facility modifies the Graph Mapping Base and it can be accessed either in the Graph Mapping Options dialog (from the pull-down menu View), or interactively by pressing the Global Font Decrease or Global Font Increase buttons on the graphics toolbar. Each click of the toolbar button changes the value of the Graph Mapping Base by 16% (i.e., one click of the button will increase the font size and decrease the mapping base by 16%). If you hold the CTRL key while clicking the Global Font Increase/Decrease buttons, the mapping base will change by 4%.
    This facility allows you to effectively increase or decrease all fonts not by changing the specific font size settings (e.g., not by changing an 8-point setting into a 12-point setting) but by globally adjusting the manner in which the logical font sizes are mapped into the plot region of the screen or the printout (see the previous topic for more information). Thus, after you press this button, a text in the graph which was set to size 8 (points) will remain set to 8, but this size (8) will now be represented by approximately 16% larger letters when displayed or printed. See also the global font scaling settings available in the Graph Mapping Options dialog (accessible from the pull-down menu View).

Whenever I print a graph, it only uses one-third of the page. How do I make the graph fill the page?
    Graphs are always printed in the current graph window proportions. This setting can be changed manually (with the mouse) by pressing the Adjustable Graph Proportions toolbar button and then adjusting the graph window proportions interactively, or by selecting Graph Mapping Options from the View pull-down menu and setting the Graph Window Proportion to Printer (or to a Custom setting). Note that the current printer setup options will determine whether the graph is printed in portrait or landscape orientation.

Printer memory requirements (for 600+ DPI printers)
    Due to the detail of STATISTICA graphs, producing hard copy output using high-resolution (e.g., 600 DPI) printers is quite memory intensive. For example, a printer needs to have at least 4 megabytes of RAM in order to reproduce all details of 3D displays (e.g., such as fonts transformed in perspective) at 600 DPI. Note that 4 megabytes is more than what is shipped as standard with some of the popular printers (e.g., HP 6P ships with 2 megabytes of RAM). If there is not enough memory in the printer, some details of 3D displays will be simplified compared to their screen appearance (e.g., the "straight" fonts will be generated). If your 600 DPI printer has less memory, set it to 300 DPI (within STATISTICA); because STATISTICA employs its own, proprietary resolution enhancement technology when generating output, the quality of printouts will be only negligibly lower (and all 3D details will appear as on screen).
    Also, please note that in STATISTICA, the default setting of the Minimum line thickness (in the Page/Output Setup dialog, select the Graphs radio button) is 2 device pixels, which is the best value for 300 DPI devices. If you are using a high-resolution printer, it is recommended to adjust this setting accordingly (i.e., to 4 or 8 pixels, on 600 and 1200 DPI printers respectively), otherwise, the output will be "too light."

GENERAL USER- INTERFACE CONVENTIONS

How to access Help for a specific dialog:
    Quick access to the Electronic Manual is provided via the Windows 95-style button at the right-hand side of the caption bar of every dialog box and floating toolbar. Simply click on this button to bring up a Help window containing the description of that dialog.

What are MicroScrolls?
    Numerical values in all dialogs may be changed by using the MicroScrolls controls. The right-mouse-button can be used to increment or decrement the next-to-last digit (e.g., clicking on the MicroScroll with the left-mouse-button increments .15 to .16, then .17, .18, etc.; if you click on the right-mouse-button, 0.15 will change to 0.25, then to .35, .45, etc.).

What are Tool Tips?
    ToolTips (Windows 95-style) are small, yellow "balloon help tips" that pop up when the mouse points to a toolbar button and help the user to quickly learn the functions of all toolbar buttons.
    You can control the display of the ToolTips in the STATISTICA Defaults: Display dialog. If a more detailed description of a button is required, you can click on the toolbar button with the left-mouse-button, but instead of releasing the mouse-button, hold it down, and a description of that button will appear in the comment area of the Status Bar.
    To obtain detailed descriptions of all the buttons on the respective toolbar, click on the Help button on the toolbar.

How to automate commonly-used procedures or repeat similar tasks in STATISTICA:
    A variety of facilities are provided in STATISTICA to automate analyses, including Auto Task Buttons, STATISTICA Command Language (SCL), STATISTICA BASIC, and two types of macros.
    Also, many modules and statistical graphs allow you to automatically repeat the same analysis for each of a series of variables or each level of a grouping variable (e.g., categorized graphs, and Quick Basic Stats).

What are Auto Task Buttons?
    This floating or docked toolbar contains user-defined buttons assigned to standard tasks, operations, or custom-defined procedures. Use the Customize... button to assign to new keys: STATISTICA Command Language (SCL) tasks, STATISTICA BASIC programs, names of data or supplementary files (e.g., graphs, reports) to be opened for quick access, or simply editable sequences of keystrokes. Pressing a button will execute a task or a series of tasks of practically unlimited complexity.

What is the Module Switcher and how does it work?
    As mentioned before, STATISTICA consists of modules, each containing a group of related procedures. When you switch modules, you can either keep STATISTICA down to one application window only, or alternatively, you can keep the previously-used modules open, because each of them can be run as a separate Windows application (see below).
    All "general-purpose" facilities (such as the data spreadsheet and all graphics procedures) are available in every module and at every point of the analysis. The Module Switcher allows you to bring up another module from a list of all modules available in your version of STATISTICA.
    The Module Switcher is similar to the Windows Task Switcher (which is called by double-clicking anywhere on the empty space on the Windows desktop) and can be invoked in a similar way, by double-clicking anywhere on the empty space within the STATISTICA application window. You can also call it by pressing the first button on every toolbar or by selecting the option Other Statistics in the pull-down menu Analysis.
    The Switcher also allows you to review brief descriptions of statistical procedures and facilities included in each module (the summary descriptions on the right side of the window are updated as you scroll through the list).
    Two ways of bringing up new modules (the general mode of the Switcher). There are two different ways in which the Module Switcher can open new modules.
    Depending on the current configuration (see the Defaults: General option from the pull-down menu Options), it can open new modules into the same or new application windows.
    (1) Single application mode. When you select the single application window mode, then switching between modules during a STATISTICA session will not open new application windows. Each new module will be opened into the same window replacing the module used before. Some users will like this "simple" mode because it keeps all analyses in a single application-window location and limits the number of programs opened on the desktop to the very minimum.
    Note that a similar effect can be achieved by pressing the End & Switch To button in the Module Switcher; the application window of the current module will close but it will not be replaced by the new one, instead, the new module will open in the "next" application window.
    (2) Multiple application mode. The main advantage of the multiple application mode is that you can run different analyses (modules) simultaneously in different, simultaneously open application windows. You can switch between the modules without closing the previous ones and take advantage of independent queues of Scrollsheets and graphs in different module windows.
    This mode has clear advantages for most types of analyses allowing the user to use (and compare results of) different analytic tools. Usually, it is not practical to keep more than four or five separate analyses open simultaneously (each with its own queue of Scrollsheets and graphs).
    The maximum number of modules which can be simultaneously opened depends on the hardware and software resources of the computer (these resources can be monitored by selecting the About... option in the Help menu of the Windows Program Manager).
    As a general rule, it is not recommended to open new applications when more than approximately 80% of Windows' resources are already used by programs which are currently running.
    The selection of the operating mode can be made using the Module Switching: Single Application Mode setting, in the Defaults: General dialog (accessible from the pull-down menu Options). If the check box is marked, then STATISTICA will run in the single-application mode.
    Customizing the list of modules in the Module Switcher. Although the list of modules in the Switcher can be scrolled, it is convenient to have the modules most commonly used in your specific work listed on the top and thus not require scrolling; the order of modules listed can be customized by pressing the Customize list... button.
    Note that the selection and order of modules listed in the Switcher is stored along with all other customizations of the STATISTICA system; thus you can maintain alternative lists for different projects. From this dialog you may also install additional modules to, or remove existing modules from the disk by clicking on the Add/Remove button (see the next topic).

How do I change the installed configuration of STATISTICA (add/remove modules from the disk)?
    You may change the STATISTICA configuration by using pressing the Add/Remove button in the Customize List of Modules dialog for the Module Switcher, or running the Reinst.exe program, which is located in your STATISTICA program directory (double-click on the icon Setup in the STATISTICA program group). This utility allows you to install any modules of STATISTICA which were not selected during a previous installation, to remove (delete) any previously installed modules, or if you are on a workstation and will use STATISTICA from a network file server, you may create your own customized working directory for STATISTICA.

How to learn what information is necessary to start an analysis (variables, grouping codes, options, etc.):
    Pressing the Help button or the F1 key will always open the relevant section of the Electronic Manual containing a comprehensive explanation of all options in the current dialog. However, all analysis definition screens in STATISTICA follow the "self-prompting" dialog conventions. The OK button is never dimmed: whenever you are not sure what to select next, simply click OK (or press ALT+O) and the program will proceed to the next logical step and ask you for specific input if it is necessary.

How to find a particular statistical procedure:
    If you are not certain where to find a particular procedure within STATISTICA, you may consult the Statistical Advisor by selecting the option Advisor from the pull-down menu Help. The program will ask you a set of simple questions about the nature of the research problem and the type of your data, then it will suggest to you the statistical procedures which appear most relevant (and tell you where to look for them in the STATISTICA system).
    A brief description of the statistical procedures and facilities included in each module is also displayed in the Module Switcher (the summary descriptions on the right side of the window are updated as you scroll through the list). You may also search for topics in the Electronic Manual by selecting the option Search for Help On... from the pull-down menu Help.

What are the specific names of program files for individual modules of STATISTICA?
    Here is an alphabetical list of program file names and applications (STATISTICA modules) which they represent.

Default file name extensions of files created by STATISTICA:
    The default file name extensions for the types of files which STATISTICA uses most often are:

File name extensions for data and graphics files exchanged by STATISTICA with other applications:

What is the difference between the "floating" Cont (i.e., continue) button on the bottom of the screen and the Continue button in the upper left corner of a Scrollsheet or graph?
    There is no difference between the functions of the two buttons as long as the number of Scrollsheets or graphs generated in one sequence by an analysis is not longer than the current length of the queue of Scrollsheets or spreadsheets (by default, the queue length is 3).
    However, when more document windows are about to be generated than can fit in the current queue, then:

• pressing the Continue button on the Scrollsheet or graph will proceed with generating the consecutive Scrollsheets and graphs which are waiting in the sequence, while
• pressing the "floating" Cont button on the bottom of the screen--which, in fact, represents the iconized output selection dialog--will bring up the dialog, breaking the sequence of consecutive Scrollsheets or graphs.

In some analyses, the Results summaries do not fit in their area (upper part) of the Results dialog. How can I review them?
    This may happen when your Windows system is configured to use a large font as the default system font (e.g., the large font is used by default by some video drivers in the 800x600 mode). This font can be changed to the standard size either by switching to an alternative video driver provided with your video card or adjusting the driver configuration settings (refer to the driver documentation provided with your video card). On some systems, the adjustment can also be made using the Windows Control Panel. If you globally adjust the size of the Windows system font, then all your Windows applications will look more standard. If you prefer to maintain the current font as the system font in other applications, then you can configure this font locally in STATISTICA. To adjust the font size and/or style, press the Dialog/Results font button in the Defaults: Display dialog (option Display... in the pull-down menu Options).

The icons on buttons in STATISTICA dialogs are smaller than the space available on the buttons. How can I adjust them?
    The reason for that is the larger default system font (as configured in your Windows system), which forces the buttons to a larger than standard size; all dialog buttons on such systems are larger. This fact is not easily noticeable in applications which (unlike STATISTICA) do not identify buttons with icons. If you prefer the appearance of STATISTICA dialogs as shown in the manual and the Technical Description brochure, adjust the current Windows system font (see the previous topic).

Why does it take such a long time to display the contents of the File pull-down menu?
    Check the file list on the bottom of the File pull-down menu to see if any files are stored an a floppy drive or a network drive that is not currently accessible. If so, edit the STATIST.INI file to remove the reference to these files.

I would like to prepare materials for a presentation and would like to make screenshots. I know STATISTICA has the Screen Catcher facility in the Edit pull-down menu. However, every time I try to use it, the dialog on the screen disappears. What happens?
    Screen Catcher (available from the Edit pull-down menu) can be used to make screenshots of scrollsheets and graphs. It is not able to make screenshots of selection dialogs or result dialogs. The combination ALT+Print Screen should be used in this case.

What files are installed to the Windows/System directory when STATISTICA is installed?
    The following files are installed to the Windows/System directory (or updated when they already exist) when STATISTICA is installed and are also used by other Windows programs:

HOW TO MAKE THE ANIMATED OVERVIEWS "RUNNABLE" FROM THE DEMO DISK:

1. Insert the STATISTICA Demo into your CD ROM drive.
2. In Windows Explorer, go to ENGLISH\MULTIMED subdirectory on the CD and copy MULTIMED.INI file to the directory on your computer where STATISTICA has been installed.
3. After copying the file, click on it with the right mouse button, go to Properties and un-check the "Read-only" attribute, then accept the Properties dialog ("OK").
4. Run STATISTICA, go to "Help\Animated Overviews >" menu, select the overview you would like to run.
5. When asked for a directory with Animated Overviews files, enter the full name of the directory on the Demo CD where they are located, i.e., [CD_drive_letter]:\MULTIMED (e.g., "E:\MULTIMED"), then click OK.
6. From this point on, STATISTICA will be able to run all available overviews from the Demo CD as long as it is in the CD ROM drive.

I am upgrading to Windows NT. Will STATISTICA run after I upgrade?
    Yes. Reinstall in the same directory to make sure that the WINHELP.INI file and registry get updated.

I am running STATISTICA at work, and it runs fine. When I took it home and installed it on my home computer, I got an error message (GPF in WINHLP32.EXE) when I tried to access the Help file. What is going on?
    Try any or all of the following: reinstall; download the latest help file from our download page; delete the .gid file associated with the help file.

After I installed STATISTICA, I can no longer double-click on my screen saver files (.scr) to launch them. How can I make this happen again?
    STATISTICA's scrollsheet files use the same extension (.scr) as Windows screen saver files do. If you change the default action of double-clicking .scr files so that it launches screen savers, you cannot double-click on scrollsheet files to automatically launch them. If you would still like to change the default action:

1. Make a backup copy of your registry.
2. Go to the Start button, choose Run, and type in regedit and click OK.
3. Double-click on HKEY_CLASSES_ROOT.
4. Scroll down until you find the .scr entry, and click on it.
5. Double-click on "Default". This will bring up an Edit String dialog box.
6. Under Value Data, enter SCRfile and click OK.
7. Close the registry, and reboot.

Can I integrate STATISTICA into other applications, such as a custom Visual Basic application?
    Yes. STATISTICA can be controlled from within other applications, such as Visual Basic, using the STATISTICA Command Language (SCL). Programs written in SCL can be executed from within Visual Basic using the SHELL command. This allows for the complete integration of STATISTICA into your custom application or other applications such as ms Excel, ms Access, or ms Word. See the STATISTICA Development Environment for information about how to manage STATISTICA data files from within Visual Basic. For details and examples please refer to Calling STATISTICA from within Visual Basic.

Can I use the libraries available in STATISTICA BASIC, such as the advanced Matrix Procedure Language, in other applications?
    Yes. STATISTICA BASIC can be called from within other applications, such as Visual Basic, using the STATISTICA Command Language (SCL). Calls to one or multiple STATISTICA BASIC programs can be written in SCL using the STBASIC command. These SCL programs can then be executed from within Visual Basic using the SHELL command. This enables you to include functions from the STATISTICA BASIC libraries, such as the advanced Matrix Procedure Language library, into your custom application or other applications such as ms Excel, ms Access, or ms Word. For details and examples please refer to Calling STATISTICA BASIC from within Visual Basic.

KEYBOARD INTERFACE

    Keyboard combinations can be used as Hot Keys in order to access various STATISTICA procedures. Hot Keys can be classified as global (available in every window) or local (for specific windows or Auto Task Buttons). You can assign a Hot Key to a macro via the Auto Task Buttons. A complete listing of the global and local Hot Keys are given below.

Global Hot Keys

Local Hot Keys

Scrollsheet Window
    The following keyboard combinations (local Hot Keys) are specific to the Scrollsheet window.

Text/output, STATISTICA BASIC, or SCL Windows
    The following keyboard combinations (local Hot Keys) are specific to the Text/output Window, STATISTICA BASIC window, or SCL window.

Graph Window
    The following keyboard combinations (local Hot Keys) are specific to the Graph window.

Graph Data Editor Window
    The following keyboard combinations (local Hot Keys) are specific to the Graph Data Editor window.

Megafile Manager Window
    The following keyboard combinations (local Hot Keys) are specific to the Megafile Manager window (see Volume III of the manual).

How do I edit subgraphs of multiple graph documents using keyboard/macros?
    To edit a subgraph (e.g., X-bar Chart from the default compound graph created by STATISTICA) via the keyboard, use the key combination: CTRL+A to select the first subgraph (i.e., "object" on the graph); then, continue to use this same key combination to toggle through the set of subgraphs until highlighting the desired graph. Finally, use the combination: ALT+Enter to open this subgraph in its own window for editing. After the editing is complete, close and update the subgraph from the File pull-down menu (ALT+F).

MOUSE CONVENTIONS

    In addition to the standard, Windows-system related applications, the mouse provides shortcuts to specific features in STATISTICA. See below for an overview of these mouse conventions and functions.

Right- and Left-Mouse-Buttons

Left-Mouse-Button
    The left-mouse-button is used for selecting (e.g., an option in a dialog), highlighting (e.g., a block of data in a spreadsheet), and dragging (e.g., dragging a block in a spreadsheet or dragging an object in a graph; see below). Double-clicking on an object with the left-mouse-button provides a direct shortcut to some of the most commonly used dialogs in STATISTICA.
    For example, double-clicking with the left-mouse-button on a spreadsheet or Scrollsheet cell, header, column name, etc., will bring up a dialog allowing you to edit the respective feature, while double-clicking with the left-mouse-button on a graph object, title, point marker, etc., will bring up the respective dialog allowing you to edit that part of the graph. In this way, you save a step whenever you want to access a commonly-used option. In addition, double-clicking on the background of STATISTICA will bring up the Module Switcher.

Right-Mouse-Button
    When you click on an object (e.g., a cell in a spreadsheet) with the right-mouse-button, a dynamic flying menu will come up.
    These flying menus allow easy access to some commonly-used options in every window in STATISTICA (e.g., graph window, spreadsheet window, Text/output Window, etc.).
    For example, if you click with the right-mouse-button on a spreadsheet cell, a flying menu of graph options, statistics options, Clipboard operations, and others will come up (see above), allowing you to quickly perform the desired operation. Clicking on the background of STATISTICA with the right-mouse-button will bring up the Module Switcher.

Selecting Items from Multiple-Selection Lists

• Click on an item to select (highlight) it. Clicking OK in the dialog will accept the selection.
• Double-click on an item to select it and accept the selection (close the dialog).
• In order to select a continuous list of items, you can (1) hold down the left-mouse-button and drag the cursor over the items that you want to highlight or (2) highlight the first item and then while holding down the SHIFT key, click on the last item that you want to select.
• In order to select a discontinuous list of items, hold down the CTRL key and click on the desired items.

Using the SHIFT and CTRL Keys with the Mouse

• You can quickly highlight a block of values in a spreadsheet or Scrollsheet by clicking on a cell to define the beginning of the block and then (e.g., after scrolling with the scrollbars) while holding down the SHIFT key, click on a cell that will define the end of the block (the SHIFT-click convention). All of the values in-between will be highlighted.
• Holding down the CTRL key when you click on an Auto Task Button will open the respective dialog for editing the contents of the button.
• Holding down the CTRL key while pressing either the Graphics window Increase Font or Decrease Font toolbar buttons will achieve a small (4% instead of the default 16%) increase or decrease in the mapping base.
• Holding down the CTRL key while pressing on the spreadsheet window Display Text Values toolbar button will switch to the value label display in the spreadsheet.
• Hold down on the SHIFT key while you are dragging the mouse in the Graphics window in order to make a square, squared rounded rectangle, circle, or quarter arc.

Other Applications

• Drag-and-Drop: Provides mouse shortcuts for moving, copying, deleting, inserting, extrapolating, etc. a block of values in the spreadsheet or Scrollsheet.
• Increase/decrease column width: Adjusts the spreadsheet or Scrollsheet column width by dragging the right column border to the desired width. The new width of the column is indicated by a dashed line.
• Split Scrolling (in spreadsheets and Scrollsheets): Splits the spreadsheet or Scrollsheet window (i.e., split scrolling) by dragging the split box (the black rectangle at the top of the vertical scrollbar or to the left of the horizontal scrollbar).
• Variable speed scrolling: Controls the speed at which you scroll (1 line at a time by moving the cursor a short distance away or one page at a time by moving the cursor further away) when you extend a block outside the spreadsheet or Scrollsheet.
• MicroScrolls: Enables you to increase or decrease the value in a numeric edit field incrementally, by either the last digit (click on the micro scroll with the left-mouse-button, e.g., 1.11, 1.12, 1.13, ...) or the last digit by a factor of 10 (click on the micro scroll with the right-mouse-button, e.g., 1.11, 1.21, 1.31, ...).
• Toolbar configuration: Double-clicking on the space between the buttons of the toolbar will toggle between four toolbar configurations (1 line, 2 lines, right & top, and left & top).
• Auto Task Buttons: Holding down the CTRL key while clicking on an Auto Task Button will open a dialog allowing you to edit the Auto Task Buttons contents.
• Auto Task Buttons toolbar: Double-clicking on the space between the buttons of the Auto Task Buttons toolbar with the left-mouse-button will resize it (alternatively, you can drag this toolbar's border with the mouse to resize it). Clicking with the right-mouse-button will bring up a flying menu of options. You can "dock" the toolbar by dragging it to an area of the STATISTICA window (above the toolbar, below the status bar, or to either side of the window) where the toolbar will change configuration and remain in that position until you drag it to a different area of the window.
• Reordering items in a list: You can reorder items in a list (e.g., Workbook, Auto Task Buttons, Multiple-Graph AutoLayout Wizard, etc.) by highlighting one or more (in a continuous or dis-continuous list) items and then moving the mouse cursor to the desired position. Clicking the left-mouse-button will then move the highlighted item(s) to the insertion point.

Graph Applications

• OLE: Links or embeds foreign document files to STATISTICA documents by dragging them with the mouse directly from File Manager (across application windows) and dropping them onto STATISTICA graphs.
• Brushing: Highlights data points from the graph by clicking on them with the brushing tool.
• Zoom in and Zoom out tools: Zooms in ("magnifies") or zooms out ("shrinks") respectively, the selected area of the graph.
• Drawing tools: Adds rectangles, ovals (or circles), polylines and free-hand drawings, arrows, etc. to a graph. Hold down on the SHIFT key while you are dragging the mouse in order to make a square, squared rounded rectangle, circle, or quarter arc.
• Resizing and moving: Resizes (drag on a "black selection square") or moves (drag the entire object) selected graph objects. For fine adjustments, see Controlling the mouse with the keyboard in graphs, below.
• Editing polyline objects: Reshapes individual segments of the polyline drawing by dragging with the mouse on either the object area black selection squares or any of the black selection squares that mark the line segments. For fine adjustments, see Controlling the mouse with the keyboard in graphs, below.
• Rotating text: You can interactively rotate custom text by selecting it in the graph and then dragging one of the object handles (small black squares) in the desired direction [the angle (in degrees) will be displayed in the Change Show window of the Graphics toolbar]. For fine adjustments (1 degree movements), see Controlling the mouse with the keyboard in graphs, below.
• Snap to Grid: Aligns drawing objects with the drawing grid, an invisible network of lines that covers the drawing area.
• Graph Mapping Base: Clicking on the Increase Font or Decrease Font toolbar buttons will change the value of the Graph Mapping Base by ±16%, respectively. Holding down the CTRL key while pressing either of these buttons will achieve a smaller (4%) increase or decrease in the mapping base.
• Controlling the mouse with the keyboard in graphs: You can also emulate the mouse with the keyboard in order to move or resize an object by selecting the object, placing the mouse over the object, and then using the keyboard cursor keys to move or resize the object. Hold down the CTRL key for fine (1 pixel only) movements and press ENTER to complete the action.

Breaking, Stopping, or Interrupting the Current Action

STATISTICS

How can I generate subgroup identifiers (codes) for my quality control charts?
    A spreadsheet formula can be used to generate subgroup identifiers. For example, the formula:
        =trunc((v0-1)/n)+1
where n is the number of times to repeat each value, can be used to generate the codes. For example, when n=4, then the following values (codes) of the grouping variable will be generated: 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3,...
    Note that the constant sample size option can also be used (instead of a grouping variable).

What if I need subgroup identifiers with repeating groups of a certain size?
    A spreadsheet formula can generate a repeating identifier variable of a certain size. For example, the formula:
        =v0-n*trunc((v0-1)/n)
where n is the maximum group size, can be used to generate a variable with codes up to that size. For example, when n=4, then the following values (codes) of the grouping variable will be generated: 1, 2, 3, 4, 1, 2, 3, 4, 1, 2, 3, 4,...

When I use the Case Weights option, I do not believe the results are correct.
    Case Weights must be integers. If the variable chosen as the weighting variable is a non-integer number it will be truncated to become an integer. The Case Weights option treats values of a selected variable as integer case multipliers when processing the data.

How do I create a control chart for a data file where each column represents an observation in a sample? For example, I have 20 samples, each of size 5, so my data file has 20 rows and 5 columns.
    You can use the Block Stats/Rows option by right mouse clicking on the desired row to compute columns of means and standard deviations. Add a column using the Vars toolbar button and fill in the sample size with a spreadsheet formula (e.g. =5). Then use the means, standard deviations, and sample sizes as the input instead of the raw data.

I am calculating a correlation matrix and am using the pairwise deletion of missing data option. Is there a way that I can find the valid n for each bivariate correlation?
    Yes, in the Pearson-Product Moment Correlation dialog in the Display box are radio buttons that allow the user to specify the content of the correlation output. Simply click on the radio button marked Corr. Matrix (highlight p & n) and the valid n that the correlation is based on using pairwise deletion of missing data will appear in the output.

I performed regression analysis on a large number of cases and created a normal probability plot, but only 2,100 cases were included in the plot. How do I create a plot that includes all of my cases?
    This limitation applies only to normal probability graphs requested from the Multiple Regression Results dialog. In order to plot all residuals, save your residuals and predicted/original variables as a data file from the Residual Analysis dialog. Then, open the data file and choose Stats 2D Graphs -- Normal Probability Plot from the Graphs pull-down menu.

In the Other Significance Tests dialog, the Print Results for each Compute option is dimmed. What do I do to make this option available?
    Select Print from the Options pull-down menu and change the Output to Printer in the Page/Output dialog.

Why do I get the message "not enough valid cases" or "less than 2 valid cases" when I try to perform an analysis (or make a graph)?
    It is likely that you have either case selection conditions or case weights turned ON. Look at the status bar in the bottom right corner of the screen to see if they are ON or OFF. If one or both are ON, they can be turned OFF from the Options pull-down menu.

When performing a factor analysis with 12000 cases, there were only 8000 factor scores listed in the results. How can I view the remaining cases?
    In STATISTICA, you can save the contents of any Scrollsheet in spreadsheet (data file) format (files with the file name extension *.sta), thus allowing you to use the contents of the output Scrollsheet as input for an analysis, or to export it to another file format. This is accomplished by selecting the 'Save as data' command available from the File... menu.
    In some analysis results dialogs, such as Multiple Regression or Factor Analysis, an option is provided to save specific results (e.g., residuals, factor scores) as a data file. These options allow the user to select variables from the data file to be saved along with the results so that the user can then use STATISTICA's vast analysis and graphics facilities for further exploration.

Why do I get different results among the ANOVA Module and the Visual General Linear Models module when I run the same type of analysis on the same dataset?
    The ANOVA module works off of variable selections differently than does the Visual General Linear Models(VGLM) module. Specifically, the ANOVA works off of the variable numbers, and VGLM (or any other Visual module) works off of the Syntax editor. When you specify the variables in the dialog box in the ANOVA module, use the number values for the variables. For instance, when you choose variable 67, named Average, it uses the 67th variable in the data file for the code.
    When the variables are specified in VGLM, the syntax editor will refer STATISTICA to the NAME of the variable. In our above example it will read as Average, NOT necessarily the 67th variable. To see this occur, if you go to the Syntax Editor you should note that it denotes variable names, not numbers. It has been created this way to make it easier for the user to customize this Editor for designs.
    Please note that this also means your data file cannot have multiple variables named with the same word. When you run a data file in VGLM and the variables are unique in names, it will call the right columns (Average is used and calls the 67th variable of the example above). If you have numerous variables with the same name, for example, Variables 70, 71, and 72 were named H1, H2, and H3; and variables 10-12 were also named H1, H2, and H3 it will use the first variables with the same name (10-12) even if you specified variables 70-72 in the variable specification dialog box. Since 70-72 were the same names as 10-12, it searched the data file for the first variable names with H1, H2, and H3. This happened to be in actuality, variables 10-12. Therefore, the results were actually using variables 10-12, not variables 70-72. This will undoubtedly give you different results that the ANOVA module that used variables 70-72.

Are there any other modules that may cause this to happen?
    Any module with the Syntax Editor (Version 5.5 modules: Visual General Linear Models, Visual Generalized Linear Models, Visual General Stepwise Regression, Visual Partial Least Squares) will have the same issue involved. Namely, it uses variable names, not numbers.

How can I stop this from happening?
    Make sure that no variables have the same name in each given data file created. With the current 8 character limitation in 5.5, this is more likely to happen. When this limitation is lifted, it will probably not have as big of an effect.

STATISTICA Development Environment

Where and how can the STATISTICA Development Environment be used?
    The STATISTICA Development Environment can be used in a variety of environments: STATISTICA BASIC, Visual C++, Visual Basic, Delphi, FORTRAN. There are two versions: 16 and 32 bit. Sample applications include third party general purpose programs for importing/exporting data, front-ends for data mining applications, various applications that interface STATISTICA with other components of enterprise-wide computing environments (e.g., one can add a front-end to a database that transfers data to STATISTICA files or use STATISTICA data files to communicate selected data to a corporate database (e.g., written in VB or C/C++)), STATISTICA BASIC programs that access data files that are currently 'not loaded' into STATISTICA (data array), such as custom applications that merge data from multiple (simultaneously open) STATISTICA data files. See also STATISTICA Development Environment and the collection of STADEV Example programs.

How can I get the STATISTICA Development Environment library?
    The necessary files are available from the directory in which STATISTICA has been installed (e.g., C:\STAT\Sta_dev). If you do not own STATISTICA or if you have an older release which does not contain these files, you may download them from this site (Download the STADEV library).

STATISTICA BASIC

How to access STATISTICA BASIC:
    STATISTICA BASIC is accessible in every module (from the pull-down menu Analysis or the floating Auto Task Buttons toolbar) and you can open multiple STATISTICA BASIC programs in one module (or in different modules).
    For example, you may have 5 or more different programs opened in their separate windows and cut and paste the code between the programs or execute them in arbitrary sequences. The BASIC comes with an integrated environment that allows you to write, edit, verify, debug (i.e., "dry run"), and execute your programs. Entering your BASIC programs is greatly simplified by using a flexible Function Wizard facility (see the next topic).

What is the Function Wizard?
    The Function Wizard is a special dialog that helps you write programs. It is accessible by pressing the Function Wizard toolbar button (or ALT+I) and allows you to review, select, edit, and insert into your program any items from structured, hierarchically arranged lists of the syntax of all BASIC functions and keywords (with all respective parameters).

What kind of variables are supported in STATISTICA BASIC?
    STATISTICA BASIC supports both temporary variables (local for the program) and the standard STATISTICA variables from the data file. The temporary variables can be either numeric (e.g., i, income, my_new_variable_with_a_long_name) or text variables (e.g., a$, b$, first_name$, last_name$), and they do not need to be declared unless they are arrays or global variables that are accessible in user-defined functions or subroutines. The data file variables can be referred by their original names, numbers, and they can be treated as vectors; data files can be treated as predefined, two-dimensional arrays ( DATA(v,c) ).

How to assign a value to a variable:
    Here are examples of simple assignment statements:
    i := 10; j := 20;
    monthly_profit := (revenues95-expenses95)/12;
    first_name$ := 'JANNICE';
    Note that in STATISTICA BASIC, the assignment operator is := and that the end of each statement is marked by a semicolon ( ; ).

How to specify a conditional instruction (if ... then):
    Here is an example of a simple conditional instruction (a 0 will be assigned to variable v2, when v1 is 0 and v2 is greater than 0):
    if (v1=0 and v2>0) then v2:=0;
    If more than one instruction is to be executed conditionally, a block of instructions can be marked with the "syntactic parentheses" BEGIN ... END; for example:
    if sum=0 then
      begin
        v7:=0;
        v8:=1;
        v9:=2+log(v10);
      end;

How to execute an expression in a loop:
    Here is an example of a simple loop:
    for i:=1 to 10 do a(i):=a(i)/1000;
Note that a semicolon marks the end of the loop.
    The following example shows a "double" (technically speaking "nested") loop. This program calculates the sum of all values in the current data file. The first loop is executed over all cases (case 1 through case NCases), the second one is executed over all variables (variable 1 through NVars) of the predefined data file array Data(NCases, NVars):
    sum := 0; {initializes a temporary variable}
    for i := 1 to NCases do
        for j := 1 to NVars do
            sum := sum + Value(Data(i,j));
    WriteLn('sum=',sum);
The end result (the value of the temporary variable sum) is then displayed using the WriteLn function.

How is STATISTICA BASIC integrated with STATISTICA data and output/input?
    Accessing data from STATISTICA system files. STATISTICA BASIC can operate directly on STATISTICA system data files, Scrollsheets, or graphs, thus, unless special applications are required, the user does not have to be concerned about file handling, input, output, etc. It supports two flexible modes of accessing raw data from STATISTICA system data files:

• spreadsheet-like Random Access Mode, where all data are loaded into memory and referenced in a predefined array DATA(case, variable) (this is the default mode), and
• data-base-like Sequential Access Mode, where data records are read sequentially, one at a time (although this can be changed with a jump command) and the entire program is automatically executed (repeated) for each record (case) of the data file.

What is the capacity of STATISTICA BASIC, can it handle large computational problems?
    STATISTICA BASIC is a powerful programming language that is well-suited to handle large computational problems: It supports local data arrays with up to 8 dimensions and there is no limit to the size of the arrays (all memory available in your system can be used, including virtual memory, e.g., correlation matrices 1,000 x 1,000 and larger can be computed [with a single function call] on most systems), so custom procedures involving operations on large multidimensional matrices can be developed (a comprehensive selection of matrix operations are also supported, see below). Matrices can be dynamically allocated or redimensioned in run time.

Does STATISTICA BASIC support statistical functions or procedures?
    STATISTICA BASIC offers numerous specialized procedures that are particularly useful for statistical data analysis and statistical database management. For example, supported procedures include correlations, crossproducts, eigenvalues, eigenvectors, a wide selection of descriptive statistics functions, custom data recoding functions, random sampling of cases, functions that create random variables that follow theoretical distributions, a wide selection of distribution functions, their inverses and integrals, specialized functions that automatically adjust for missing data, conditional deletion of cases, and many others.
    Also, a variety of statistical procedures are available in the Matrix Library (see below) of STATISTICA may be applied to the current data file or to specific variables in the current data file.

Does STATISTICA BASIC support graphics?
    STATISTICA BASIC is a powerful graphics development tool. It offers access to virtually all STATISTICA graphics procedures. Creating even complex graphs with STATISTICA BASIC is surprisingly easy. A Wizard is provided to speed up entry of all commands (including all graphics functions). For example, only one line of code is all that is necessary to create a graph (e.g., from a particular combination of values in the data file or any data computed in your program).
    The following (complete) STATISTICA BASIC program will create a scatterplot from all cases (NCases) of the first two variables (v1 and v2) in the current dataset.
    NewGraph (SCATTERPLOT, 'Title', 'x', 'y', NCases, v1, v2);
    The following modification of that program will label the axes of the graph with the names of the respective variables (these names are placed in string variables x$ and y$, then used as axis labels).
    x$ := VarName(1);
    y$ := VarName(2);
    NewGraph (SCATTERPLOT, 'Title', x$, y$, NCases, v1, v2);
    Applications. The applications of the graphics functions of the BASIC are virtually countless and range from simple sequential line graphs (e.g., created for every case in a data file and overlaid in one display) to new, complex types of specialized multi-graphics displays, technical drawings, and diagrams linked to your data (this also includes all STATISTICA's complex types of graphs, such as categorized 3D and ternary graphs, creating dynamic true-3D objects in perspective [i.e., true 3D-drawing, as shown in the STATISTICA BASIC graphs shown below] and other advanced functions).
    The BASIC can also be used to automate routine sets of modifications or customizations of existing graphs (e.g., you can design a library of your own menu-driven graph customization procedures). STATISTICA BASIC can create complex compound graphics documents with OLE links (including nested documents), diagrams related to data (that can be updated/redrawn by pressing a button), and many other types of displays.
    Graphics output. The graphics output can be combined with numeric or text output and sent to the report editor to produce customized reports. The graphs can also be saved or printed directly.
    User-interface for BASIC graphics procedures. Your new, custom-designed graphics procedures can be added to the STATISTICA system by assigning them to buttons on the floating Auto Task Buttons toolbars. Simple to use, predefined dialogs can be set up (with simple function calls, see below) to produce customized "front ends" for these new procedures (e.g., prompting the user to enter the necessary parameters, text of titles, select specific options, or any other type of input).

Does STATISTICA BASIC support matrix operations?
    STATISTICA BASIC includes an integrated, powerful Matrix Procedure Language allowing you to develop your own statistical and other computational programs by writing compact code referencing a comprehensive library of highly optimized matrix procedures that can operate on arrays of values of practically unlimited size (the arrays can be dynamically allocated or redimensioned in run time, see above).
    The supported operations include a complete matrix algebra language and a comprehensive selection of specialized procedures that are particularly useful in the development of custom statistical applications:

• functions for computing the determinant, inverse, eigenvalues, eigenvectors, sweeping, Gram-Schmidt orthonormalization, Cholesky decomposition, singular values, singular value decomposition, matrix rank, solutions to linear equations, general pseudo-inverse;
• functions for computing correlations and covariances (with or without the intercept);
• functions for extracting specific columns, rows, submatrices, a diagonal, transposing, and combining matrices;
• addition, subtraction, multiplication, division by scalar (for each element of the matrix);
• addition, subtraction, multiplication of matrices;
• elementwise relational operators (<, <=, >, >=, =, not =);
• elementwise logical operators (not, and, or);
• tests for all elements 0, any element 0 ;
• options to read matrices from datasets, Scrollsheets, ASCII text files, Scrollsheet files, save as datasets, Scrollsheets, ASCII files.

Are there example STATISTICA BASIC programs included in the package?
    A comprehensive library of example STATISTICA BASIC programs and applications (a complete source code ready to run) is provided with your program (and on this web site) to illustrate advanced applications. This includes (among others):

• a General Multiple Regression program (with user-interface, graphs, etc.) that supports regression models with practically unlimited numbers of variables;
• a General Weighted Least Squares Regression program (see the illustration, below):
• a Two-stage Least Squares Regression program;
• general Box-Cox and Box-Tidwell transformation procedures with graphs;
• a General Factor Analysis program that can handle models with practically unlimited numbers of variables.

What kind of output can I create with STATISTICA BASIC?
    The output from STATISTICA BASIC can be handled in a variety of ways. First, the values (numeric or text) can be written directly to the current STATISTICA data file (e.g., when you need to perform data transformations or cleaning). It can also be directed to the printer, disk files, interactive editor dialogs, or the report window (where it can combine graphs with text or tables in a single document). The syntax of output control in STATISTICA BASIC is very simple; for example, many advanced output objects can be created with just one line of code. Also, text or tabular output can be sent to new Scrollsheets (see the next topic).

Can I create full-featured Scrollsheets in STATISTICA BASIC?
    In addition to the various report, external file, or data file output options (see the previous topic), you can create new Scrollsheets or cascades of Scrollsheets (directly from within your BASIC program) and display your tabular or text results in a way that is most efficient to review. Creating a simple output Scrollsheet takes just one line of code, and as a result. The following program will compute a correlation matrix and display it (and then print) in a Scrollsheet:
    ReDim corr(NVars, Nvars);
    {declares a square array corr of size equal to
    the number of variables in the current data file}
    MatrixCorrelations(Data, 1, corr);
    {computes correlations from data in the array corr}
    name$ := VarName(1);
    {places 1st variable name in string variable name$}
    for i := 2 to NVars do
    {beginning of a loop: FOR..TO..DO..}
    name$ := name$ + '|' + VarName(i);
    {builds a delimited string containing variable names}
    scr := NewScrollsheet(NVars, Nvars, corr, 'Correlations', name$, name$);
    {displays the correlation matrix in a new Scrollsheet}
    PrintScrollsheet(scr);
    {Prints the Scrollsheet to the printer}
    Advantages of Scrollsheet output from BASIC programs. These simple Scrollsheet BASIC functions generate "tools" to review tables and values that are incomparably more efficient and user-friendly than traditional text-output editors. For example, the value display format, column width, etc., can be interactively customized by the user of your STATISTICA BASIC programs, edited (including Drag-and-Drop, AutoFill, etc.), printed as tables with high-resolution grid lines, selectively copied, exported, etc.; the values can also be interactively visualized in graphs of any type supported in STATISTICA.
    Customized Scrollsheets. If you need to create specific, highly customized Scrollsheets that look like those created by the standard STATISTICA procedures (with titles, custom formatting, fonts, colors, highlighting, etc.), a comprehensive set of functions is provided and can be used in your BASIC programs.

Can I access data from existing Scrollsheets in STATISTICA BASIC and do "follow-up analyses"?
    Yes, you can also easily write procedures to manage existing Scrollsheets. This includes not only global operations (e.g., save, open, print Scrollsheets), but Scrollsheets can also be used for input. Specifically, you can read data from existing Scrollsheets created by other STATISTICA procedures, and thus your BASIC programs can be designed to "continue" or "follow up" on analyses completed by other procedures of STATISTICA. For example, you can write procedures to further analyze frequency data from a crosstabulation or from complex sequences of means displayed in a multiple-classification table; you can analyze output from Monte Carlo procedures, or just rearrange the layout of an existing Scrollsheet (e.g., by reading its contents into a BASIC data array and then writing them to new, custom-designed Scrollsheets); you can merge or split Scrollsheets, save them as STATISTICA datasets, Scrollsheet files, or ASCII files. Like all other functions in STATISTICA BASIC, all Scrollsheet functions are easy to use because they can be inserted with the help of the Wizard (see above).

Can I create custom dialogs and other interactive user-input controls in STATISTICA BASIC?
    Simple to use and flexible facilities can be called using STATISTICA BASIC with a single function call to accept the user input. For example, a statement as simple as Read(A) will display a dialog box and prompt the user of your BASIC program to enter a value for variable A; the statement PauseAndEdit will display an interactively customizable dialog where you can display and/or edit (in run time) current values of any variable in the program including values from all arrays (an invaluable aid for debugging your programs, but also a perfect tool to be incorporated in your custom BASIC applications). A library of flexible functions that generate the most commonly used types of dialogs is available.
    For example, with a single line of code (that can be inserted with a single click of the mouse via the Wizard), you can create dialogs for displaying messages, warnings, Yes/No/Cancel selections, etc., dialogs with buttons, as well as more complex custom-input dialogs (e.g., single-selection or multiple-selection list boxes, data entry dialog boxes with multiple fields for numeric and/or text values, etc.).
    A set of simple to use functions is also provided to call the standard STATISTICA variable selection dialogs (single or multiple list dialogs are supported, with built-in spread/shrink, zoom, review values, and other functions) as well as dialogs to select file names for opening or saving.
    The selection of those standard Windows resources and STATISTICA-specific user interface tools callable from the BASIC is quite comprehensive and sufficient in most standard applications. However, if you have the need for more specialized or customized dialogs, floating toolbars, palettes, etc., you can define them using standard Windows development tools and call them from your STATISTICA BASIC program as an external DLL function (e.g., see the custom data entry dialog box shown in the illustration on the right; STATISTICA BASIC supports external DLLs, see the next topic).

Can I expand my STATISTICA BASIC programs by calling external DLLs (procedures written in other programming languages)?
    Your STATISTICA BASIC applications can also be further extended by calling procedures written in other programming languages. STATISTICA BASIC can call external programs via a flexible interface to external DLLs (Dynamic Link Libraries), thus your STATISTICA BASIC programs can serve as an interface between existing applications written in other programming languages (e.g., commercial or public domain procedures) and your STATISTICA data files, graphs, and reports. Also, STATISTICA BASIC can be used to develop STATISTICA-compatible front ends for other applications (e.g., see the example with ms Organization Chart in the illustration above).

Can I use STATISTICA BASIC to develop "extensions" of STATISTICA and my own "new modules"?
    Your STATISTICA BASIC applications, such as custom data transformations, computational/analytic procedures, interfaces to other programs, or new, user-designed graphs can be permanently added to your STATISTICA system and made easily accessible by assigning them to the Auto Task Buttons on the floating or docked toolbars. Libraries of your programs can be easily organized using local or global toolbars. STATISTICA BASIC can also be used to develop your own custom-designed "STATISTICA modules" or "STATISTICA applications" based on the STATISTICA File Server foundation (included in the package).

What is the difference between Sequential and Randomaccess modes of STATISTICA BASIC?
    Sequential mode treats the STATISTICA dataset like a database, executing the entire STATISTICA BASIC program once for each record (case) of the database. The consecutive order of processing records can only be changed by using the Jump(x) command.
    Randomaccess mode treats the STATISTICA dataset like a spreadsheet, with all values contained in the predefined array DATA(i,j), where i represents the variable (column) number and j the case (row) number within the spreadsheet.
    In Randomaccess mode, the STATISTICA BASIC program is executed once and any data cell may be accessed by referencing both the variable and case numbers. Review the respective topics in the Electronic Manual for a discussion of the specific advantages and applications of the two modes.

How can I use my Quick MML programs from STATISTICA version 4.5 in STATISTICA BASIC?
    Because Quick MML operates only in Sequential mode, and STATISTICA BASIC can run in either Randomaccess (default) or Sequential mode (see the previous topic), you need to specify the mode (enter the word "Sequential") in the first line of your Quick MML program.

How to set (or leave unchanged) default values of function parameters:
    Many functions in STATISTICA BASIC have multiple parameters; to simplify programming, STATISTICA BASIC allows you to ignore those parameters which are not absolutely necessary. You can do this in one of two ways:
    (1) Instead of specifying the value of a parameter, place a ? before any word in place of the parameter (for text parameters, a $ must be used at the end of the word). Note that if you insert the function from the Function Wizard dialog, this is how the parameter keywords will appear until you change them to the desired values. (These can even be abbreviated as ? for numeric parameters or ?$ for text parameters.) For example, the following program will leave the titles in the resulting scatterplot unchanged and place the default arrow type, style, size, and angle in the specified position on the graph.
    my_graph := NewGraph(SCATTERPLOT, ?Top_Title$, ?Left_Title$,
        NCases, v5, v10);
    GraphDrawArrow(graph, ?ArrowType, ?Size, ?Angle, ?Style, 43, 55, 66, 59);
    (2) Instead of specifying the value of a parameter, use the keyword UNUSED (or UNUSED$ for text parameters) if you do not want to change the parameter and accept the default settings. For example, the above program could be written as:
    my_graph := NewGraph(SCATTERPLOT, UNUSED$, UNUSED$, UNUSED$,
        NCases, v5, v10);
    GraphDrawArrow(my_graph, UNUSED, UNUSED, UNUSED, UNUSED, 43, 55, 66, 59);
    in order to achieve the same results.

How do I remove cases that have values greater than 2.5 (or some other multiple) of standard deviation for a variable (or set of variables)?
    You can write a STATISTICA BASIC program, using if then statements to find out which values are outside of +/-2.5 standard deviations and use the Delete command in Sequential Mode to delete those cases.

How do I use Windows API calls in my STATISTICA BASIC programs?
    STATISTICA BASIC can be used to call Windows API functions. In order to do this, you must first Declare the function at the beginning of the STATISTICA BASIC programs (before any instructions are executed), and then later in the program, you can call the function. For details and examples please refer to the How to use Windows API calls in STATISTICA BASIC section in the Electronic Manual or the WinApi.stb example on this site. For a list of all the STATISTICA BASIC example programs available on this site, see the STATISTICA BASIC Programs (User-exchange Forum).

Can I execute external applications from within STATISTICA BASIC?
    Yes. STATISTICA BASIC can be used to execute other applications (.exe files) and, if desired, open specific files in that application. In order to do this, you must first Declare the ShellExecute function at the beginning of your STATISTICA BASIC program (before any instructions are executed), and then later in the program, you can call the function. For details and examples please refer to the How to execute applications (.exe files) from STATISTICA BASIC section in the Electronic Manual or the CallExe.stb example on this site. For a list of all the STATISTICA BASIC example programs available on this site, see the STATISTICA BASIC Programs (User-exchange Forum).

2300 East 14th Street, Tulsa, OK 74104
Phone: (918) 749-1119; Fax: (918) 749-2217
e-mail: info@statsoft.com

©Copyright StatSoft, Inc., 1984-2004.
StatSoft, StatSoft logo, STATISTICA, SEWSS, SEDAS, Data Miner, SEPATH and GTrees are trademarks of StatSoft, Inc.