Monthly Archives: April 2013
STATISTICA Multivariate Statistical Process Control
STATISTICA Multivariate Statistical Process Control (MSPC) is a complete solution for multivariate statistical process control, deployed within a scalable, secure analytics software platform.
Modern automated production processes typically measure large numbers of variables. Examples of products include biochemicals, cement, fertilizers, food, paint, perfume, pharmaceuticals, petroleum products, polymers, pulp, and semiconductors.
Common goals for the production process is to reduce product variability and increase quality. Finding problems sooner rather then later has the potential to save money.
But standard quality control charting techniques (e.g., Shewhart charts, Xbar and R charts, etc.) are applicable only to single variables. Therefore, when applied to modern production processes with hundreds of important variables that need to be monitored, the criteria typically applied to univariate charts will lead to a large number of false alarms, and in many cases nearly constant, perpetual, alarms.
To rectify these shortcomings, methods were developed to monitor simultaneously multiple variables, using multivariate statistical procedures.
STATISTICA‘s MSPC capabilities allow you to:
 Apply univariate and multivariate statistical methods for quality control, predictive modeling, and data reduction to complex manufacturing processes
 Determine the most critical process, raw materials, and environment factors and their optimal settings for delivering products of the highest quality
 Monitor the process characteristics interactively or automatically during production stages, rather than waiting for final testing
 Build, evaluate, and deploy predictive models based on the known outcomes from historical data
MSPC Features
 Offline Analyses vs Online Analyses
 MSPC Deployment
 Principal Components Analysis (PCA)
 Partial Least Squares (PLS)
 BatchWise MultiWay Partial Least Squares (BMPLS)
 TimeWise MultiWay Principal Component Analysis (TMPCA) and TimeWise MultiWay Partial Least Squares (TMPLS)
Process Analytical Technology
The goal of PAT is to understand and control the manufacturing process, which is consistent with our current drug quality system: quality cannot be tested into products; it should be builtin or should be by design. FDA – Process Analytical Technology 
STATISTICA MultiStream for Pharmaceutical Industries is the solution package for PAT applications. Validation packages (IQ/OQ/PQ) and Validation Services are available for purchase.
Details
STATISTICA can support different modes for employing MSPC techniques.
Offline Analyses
 Historical analysis, data exploration, data visualization, predictive model building and evaluation, model deployment to monitoring server
Online Analyses
 Interactive Monitoring with Dashboard summary displays and Automaticupdating results
 Automated Monitoring with rules, alarm events, and configurable actions
MSPC Deployment
Deployment enables you to apply existing models created from STATISTICA MSPC, to new data in order to make further predictions. You can save models in C\C++, Visual Basic, and PMML formats. But MSPC will only accept PMML for deployment.
Principal Components Analysis (PCA)
The aim of Principal Components Analysis (PCA) is to reduce the dimensionality of a set of variables while trying to preserve as much information contained in the data as possible.
Equally important applications of PCA include data diagnostics, both on observation and variable levels. The observation level helps us to detect outliers, while the variable level provides us with insight of how the variables contribute to the observations and relate (correlate) to one another.
These diagnostic features of STATISTICA PCA are particularly useful for process monitoring and quality control as they provide us with effective and convenient analytic and graphic tools for detecting abnormalities that may rise during the development phase of a product. PCA data diagnostics also play an important role in batch processing where the quality of the end product can only be ensured through constant monitoring during its production phase.
Nonlinear Iterative Partial Least Squares (NIPALS) algorithm can be used within PCA.
Partial Least Squares (PLS)
Partial Least Squares (PLS) (also known as Projection to Latent Structure) is a popular method for modeling industrial applications. It was developed by Wold in the 1960s as an economic technique, but soon its usefulness was recognized by many areas of science and applications including Multivariate Statistical Process Control (MSPC) in general and chemical engineering in particular.
It many ways, PLS can be regarded as a substitute for the method of multiple regression, especially when the number of predictor variables is large. In such cases with regression, there is seldom enough data to construct a reliable model that can be used for predicting the dependent data Y from the predictor variables X. Instead, we get a model that can perfectly fit the training data while performing poorly on unseen samples. This problem is known as overfitting (Bishop 1995).
PLS alleviates this problem by adopting the “often correct” assumption that, although there might be a large number of predictor variables, the data may actually be much simpler and can be modeled with the aid of just a handful of components (also known as latent).
Nonlinear Iterative Partial Least Squares (NIPALS) algorithm can be used within PLS.
BatchWise MultiWay Partial Least Squares (BMPLS)
Although the PLS method is extremely useful for tackling MSPC problems, it is strictly applicable to 2dimensional problems. Thus, before analyzing 3dimensional batch data it must be transformed.
This can be achieved using the method of unfolding batchwise. The 3dimensional matrix is unfolded in the direction of the batches.
TimeWise MultiWay Principal Component Analysis (TMPCA) and TimeWise MultiWay Partial Least Squares (TMPLS)
Batch processes are by nature time based. It is not only the trajectory of the batch variables that vary in time but also the correlation among them. Therefore, any monitoring system should implicitly include this dynamic time dependency. That is why timewise unfolding is particularly suited for online batch monitoring since it preserves the time dimension that is inherent in the data set.
As a result, PCA and PLS models built on timewise unfolding are particularly sensitive, not only to the quality of a batch as a whole but also to the time dependent conditions under which the batch was evolved. Thus, they are better suited for online monitoring.
System Requirements
STATISTICA Multivariate Statistical Process Control is compatible with Windows XP, Windows Vista, and Windows 7.
Minimum System Requirements
 Operating System: Windows XP or above
 RAM: 256 MB
 Processor Speed: 500 MHz
Recommended System Requirements
 Operating System: Windows XP or above
 RAM: 1 GB
 Processor Speed: 2.0 GHz
Native 64bit versions and highly optimized multiprocessor versions are available.
STATISTICA Power Analysis and Interval Estimation
Using STATISTICA Power Analysis and Interval Estimation in planning and analyzing your research, you can always be confident that you are using your resources most efficiently. Nothing is more disappointing than realizing that your research findings lack precision because your sample size was too small. On the other hand, using a sample size that is too large could be a significant waste of time and resources.
STATISTICA Power Analysis and Interval Estimation will help you find the ideal sample size and enrich your research with a variety of tools for estimating confidence intervals.
Read more about STATISTICA Power Analysis and Interval Estimation:
 Advantages
 Power Calculation
 Sample Size Calculation
 Interval Estimation
 Probability Distributions
 List of Tests
 Example Application
Details
Advantages
Some of the advantages of STATISTICA Power Analysis and Interval Estimation are:
 Precise and fast computational routines, which maintain their accuracy across a broad range of parameters
 Presentationquality, automaticallyscaled graphs of power vs. sample size, power vs. effect size, and power vs. alpha
 Protocol statements describing calculations in a form that can be transferred directly to a text document
Power Calculation
Power Calculation allows you to calculate statistical power for a given analysis type (see List of Tests below), and to produce graphs of power as a function of various quantities that affect power in practice, such as effect size, type I error rate, and sample size.
Sample Size Calculation
Sample Size Calculation allows you to calculate, for a given analysis type (see List of Tests below), the sample size required to attain a given level of power, and to generate plots of required sample size as a function of required power, type I error rate, and effect size.
Interval Estimation
Interval Estimation allows you to calculate, for a given analysis type (see List of Tests below), specialized confidence intervals not generally available in generalpurpose statistical packages. These confidence intervals are distinguished in some cases by the fact that they refer to standardized effects, and in others by the fact that they are exact confidence intervals in situations where only approximate techniques have generally been available.
STATISTICA Power Analysis and Interval Estimation is unique among programs of its type in that it calculates confidence intervals for a number of important statistical quantities such as standardized effect size (in ttests and ANOVA), the correlation coefficient, the squared multiple correlation, the sample proportion, and the difference between proportions (either independent or dependent samples).
These capabilities, in turn, may be used to construct confidence intervals on quantities such as power and sample size, allowing the user to utilize the data from one study to construct an exact confidence interval on the sample size required for another study.
Probability Distributions
Probability Distributions allows you to perform a variety of calculations on probability distributions that are of special value in performing power and sample size calculations.
The routines are distinguished by their high level of accuracy, and the wide range of parameter values for which they will perform calculations. The noncentral distributions are also distinguished by the ability to calculate a noncentrality parameter that places a given observation at a given percentage point in the noncentral distribution. The ability to perform this calculation is essential to the technique of “noncentrality interval estimation”
These routines, which include the noncentral t, noncentral F, noncentral chisquare, binomial, exact distribution of the correlation coefficient, and the exact distribution of the squared multiple correlation coefficient, are characterized by their ability to solve for an unknown parameter, and for their ability to handle “nonnull” cases.
For example, not only can the distribution routine for the Pearson correlation calculate p as a function of r and N for rho=0, it can also perform the calculation for other values of rho. Moreover, it can solve for the exact value of rho that places an observed r at a particular percentage point, for any given N.
List of Tests
STATISTICA Power Analysis and Interval Estimation calculates power as a function of sample size, effect size, and Type I error rate for the tests listed below:
 1sample ttest
 2sample independent sample ttest
 2sample dependent sample ttest
 Planned contrasts
 1way ANOVA (fixed and random effects)
 2way ANOVA
 Chisquare test on a single variance
 Ftest on 2 variances
 Ztest (or chisquare test) on a single proportion
 Ztest on 2 independent proportions
 Mcnemar’s test on 2 dependent proportions
 Ftest of significance in multiple regression
 ttest for significance of a single correlation
 Ztest for comparing 2 independent correlations
 Logrank test in survival analysis
 Test of equal exponential survival, with accrual period
 Test of equal exponential survival, with accrual period and dropouts
 Chisquare test of significance in structural equation modeling
 Tests of “close fit” in structural equation modeling confirmatory factor analysis
Example Application
Suppose you are planning a 1Way ANOVA to study the effect of a drug.
Prior to planning the study, you find that there has been a similar study previously. This particular study had 4 groups, with N = 50 subjects per group, and obtained an Fstatistic of 15.4.
From this information, as a first step you can (a) gauge the population effect size with an exact confidence interval, and (b) use this information to set a lower bound to appropriate sample size in your study.
Simply enter the data into a convenient dialog, and results are immediately available.
In this case, we discover that a 90% exact confidence interval on the rootmeansquare standardized effect (RmsSE) ranges from about .398 to .686. With effects this strong, it is not surprising that the 90% post hoc confidence interval for power ranges from .989 to almost 1. We can use this information to construct a confidence interval on the actual N needed to achieve a power goal (in this case, .90). This confidence interval ranges from 12 to 31. So, based on the information in the study, we are 90% confident that a sample size no greater than 31 would have been adequate to produce a power of .90. 

Turning to our own study, suppose we examine the relationship between power and effect size for a sample size of 31. The first graph shows quite clearly that as long as the effect size for our drug is in the range of the confidence interval for the previous study, our power will be quite high, should the actual effect size for our drug be on the order of .25, power will be inadequate.


If, on the other hand, we use a sample size comparable to the previous study (i.e., 50 per group) we discover that power will remain quite reasonable, even for effects on the order of .28.
With STATISTICA Power Analysis and Interval Estimation, this entire analysis runs in just a minute or two. 
System Requirements
STATISTICA Power Analysis is compatible with Windows XP, Windows Vista, and Windows 7.
Minimum System Requirements
 Operating System: Windows XP or above
 RAM: 256 MB
 Processor Speed: 500 MHz
Recommended System Requirements
 Operating System: Windows XP or above
 RAM: 1 GB
 Processor Speed: 2.0 GHz
Native 64bit versions and highly optimized multiprocessor versions are available.
STATISTICA Financial Analytics Solutions Continue to Earn High Marks in European Markets
TULSA, OK, USA [April 8, 2013] – StatSoft, Inc., announces that two of its analytics software solutions, STATISTICA Scorecard and STATISTICA Decisioning Platform ®, have earned top recognition in independent European competitions among financial software vendors.
STATISTICA Scorecard recently earned “Best of 2013” recognition at the IT Innovation Awards organized by Initiative Mittelstand, the “Small and MediumSized Enterprise (SME) Initiative” founded by Huber Verlag, a leading publisher in Germany.
Additionally, STATISTICA Decisioning Platform earned a recommendation award in the recent “Hit of the Year for Financial Institutions 2013” competition, organized by Gazeta Banking, Poland’s largest banking magazine.
The “BEST OF 2013” award:
StatSoft (Europe) GmbH, a StatSoft regional office in Hamburg, submitted STATISTICA Scorecard to the annual IT Innovation Awards, which honors those products and services with high economic, social, technological, and/or ecological value. Now in its tenth year, the competition saw more than 60 applicants in this category. The independent jury consisted of nearly 100 scientists, professors, IT experts, and specialized journalists who judged applicants based on level of innovation, recognizable benefits, and suitability for SMEs. STATISTICA Scorecard was awarded the title “Best of 2013” in the “Finance” category.
Hundreds of guests from politics, business, and the ICT industry attended the awards ceremony, which was held March 5 at the CeBIT 2013 Global Conference in Hannover, Germany.
This marks the second year that StatSoft’s solutions have received a top rating in the Initiative Mittelstand competition.
The “HIT OF THE YEAR” award:
In Poland, StatSoft Polska Sp. z o.o., a StatSoft regional office based in Krakow, accepted Decisioning Platform’s “Hit of the Year” recommendation award at the 11th Annual Technology Gala, hosted at the Copernicus Science Center in Warsaw. This annual competition evaluates vendors who have proposed innovative IT solutions and services designed to meet the needs of the modern financial sector in the coming year.
STATISTICA Decisioning Platform received its recommendation award as a potential market hit in the “Solutions” category, which specifically evaluates submissions based on IT and telecommunication solutions that aim to improve the performance of financial institutions.
This award represents the third consecutive year that StatSoft’s solutions have been recognized in the “Hit of the Year” competition. StatSoft Polska accepted an award in each of the 2011 and 2012 competitions for its application of STATISTICA Scorecard with financial sector customers.
Gazeta Banking’s Technology Gala was held March 13 and attracted more than two hundred highranking executives from the world of finance and IT companies. Judging was conducted by a panel comprising independent experts from the fields of economics, banking, finance, and information technology.
Big Data: Too big to fail?
 From Click Z, May 2012: Big Data Is Never Too Big When You Can Act On It
[Insight is great as far as it goes, but Big Data projects will fail without conversion of insights into actions.]  From Wall Street Journal, March 2013: Big Data, Big Blunders
[Five Big Data mistakes that companies commonly make, and expert tips on how to avoid them.]  From earlier this week at Tech World News, April 2013: 5 strategic tips for avoiding a big data bust
[More tips to increase the chances your Big Data projects do not end in failure.]
TAURON Power Company Achieves Success with STATISTICA
European electric power company has announced successful results of STATISTICA integration for purposes of data mining, forecasting, and consulting.
Tauron Dystrybucja S.A. in Krakow, Poland, selected STATISTICA Data Miner to forecast electric energy supplies while accounting for such details as territorial division, consumer segmentation, voltage levels, and contractual amounts. The company’s Director of Distribution Services Sales, Grzegorz Marek, expressed pleasure not only with Data Miner’s accuracy, but also with StatSoft’s fourday dedicated training. “Due to our substantial expert knowledge of forecasting issues, and owing to StatSoft’s supplemental expertise,” says Marek, “we are now able to prepare forecasts whose accuracy surpasses that of the Microsoft Office tools which we had been using.” READ MORE >>
Visual Introduction to STATISTICA
Efficient User Interface, Quick Templates for Novices

Make “Your Own STATISTICA” – Put a Simple Face on Advanced Functionality

Unmatched Quality of Presentation and Analytic Graphs

Full WebEnablement Analytics: WebSTATISTICA

Manage Output More Efficiently

Virtually No System Limits

. 
Builtin Visual Basic, HTML Support, and .NET Compatibility Fully Programmable System – Automate Virtually All Analyses


Powerful Querying, Rapid Data Access, and ETL

Data Visualization
Overview
Graphics facilities in STATISTICA combine an extremely wide selection of scientific and technical charts (featuring builtin analytic facilities) with customization, drawing, and multigraphics management capabilities that are usually found only in designated presentation graphics and drawing programs. STATISTICA offers hundreds of types of 2 and 3dimensional graphical displays, including 2 and 3dimensional ternary graphs, special 4dimensional graphs, multidimensional graphs, categorized multigraphs, matrices of graphs, icons, tessellations, spectral 2 and 3dimensional graphs, compound graphs, and many other specialized procedures. Also, flexible and very easy to use facilities are provided to custom design completely new types of graphs and add them permanently to the menu or floating toolbars.
STATISITCA graph display technology automatically detects and takes advantage of highperformance hardware acceleration, which is now available not only in the highend, but also in many midrange video display controllers available in both desktop and laptop computer workstations. The resulting output is generated quickly and also supports more advanced smoothing and gradient display options. STATISTICA graphs feature interactive graphic controls which enable you to interactively adjust display features.
There are various methods to request STATISTICA Graphs. You could say that these methods represent different types of “interfaces” between numbers and graphs.
For example, the numbers represented in a pie chart can simply depict values of a spreadsheet column (e.g., variable Sales) in the consecutive cases of the spreadsheet (e.g., cases labeled: Year 2002, Year 2003, Year 2004, …, etc.). The numbers in a similar pie chart, however, can also represent results of some calculations. For example, the slices of the pie can represent relative frequencies of observations that belong to certain categories calculated by one of the histogram or frequency categorization procedures (e.g., numbers of years when the Sales were below $10 million, between $10 and $20 million, and above $20 million).
Regardless of the method used to create a graph (i.e., regardless of where the numbers represented in the graph were obtained or how they were calculated), all STATISTICA Graph customization and multigraphics management facilities can be used to change the appearance of the graph or integrate it with other graphs or documents.
Also, all integrated analytic facilities that are accessible from within graphs in STATISTICA (such as function fitting, smoothing, rotation, brushing, analytical zooming, etc.) are available and can be applied to the graph regardless of the source of the numbers in the graph or the method that was used to create it. 
The graph editing facilities offered in STATISTICA enable you to create not only highly customized scientific and technical publicationready displays: 
and precise drawings: 
but also presentationquality diagrams, posters, business charts, and other displays: that are designed to communicate information in an effective and attractive manner. 
Graphs that are saved into files or that in any other way have been temporarily detached from the STATISTICA application (e.g., copied to the Clipboard or linked to a document in another application) are complete “objects” (technically speaking, ActiveX objects) that contain not only all customization features and other embedded objects, but also all data that are necessary to continue editing all aspects of the display or the analysis of its contents (fitting, smoothing, etc.).
Because STATISTICA Graphs are ActiveX objects, they can easily be linked to or embedded into other compatible documents (e.g., Excel or Word documents), where they can be inplace edited by doubleclicking on them. STATISTICA Graphs are also ActiveX containers and, therefore, can contain a wide variety of embedded or linked documents such as Visio drawings, Adobe illustrations, Excel spreadsheets, or Word documents. Moreover, STATISTICA supports hierarchies of embedded objects up to four levels, which means that it can manage “documents, containing documents, containing documents, which contain documents.”
General Types of Graphs vs. Graphs Integrated with Statistical Procedures
This section describes the general types of graphs that are available at any point of your analysis and that can be applied to arbitrary selections of data, including:
 raw data (or any subsets of raw data, specified online using flexible selection conditions);
 any output from analyses (or any subsets of the output selected in the results spreadsheets);
 values calculated in STATISTICA Visual Basic;
 any combinations of the previous three types of data.
Note that one of the unique features of the graphics facilities in STATISTICA is that all numerical values (and their text descriptions) as well as all possible combinations of those values can be visualized using all graphical procedures available in the system.
In addition to those general types of graphs (described in this section), there are hundreds of more specialized graphs that are integrated with specific statistical procedures and available either from the respective output selection dialogs or shortcut menus in the results spreadsheets. Some of those specialized types of graphs are listed in the respective topics of the Statistics section. Finally, all graphics options and procedures can also be used in your STATISTICA Visual Basic programs.
General Graphics Options
The selection of types, styles, and options of graphs that can be produced by STATISTICA exceeds by far the limits of what would be reasonable to describe in detail in this overview of features. There are hundreds of predefined and specifically preconfigured graphs that are accessible from all statistics, graphics, and shortcut menus (and from the toolbars). In addition, there are virtually countless graphs that can be custom defined by the user, and those graphs represent any combinations of userselected parts of the numeric output and raw data.
Moreover, each existing graph (either predefined or customdefined) can be treated as a starting point for unlimited customizations that can also involve changing of graph types. These customizations can go far beyond the interactive changing of attributes of all graph components and drawings. New series of data can be added or merged into an existing graph, practically all structural aspects of the graph can be redefined and customized, and new STATISTICA graphs (and/or artwork from other applications) can be dynamically linked to (or statically embedded into) the current display. Foreign files can simply be dragged onto STATISTICA graphs directly from Windows Explorer (across application windows). STATISTICA graphs can serve as “containers” for ActiveX/OLE compliant or incompatible objects pasted from other applications.
Graph (“SelfContained”) Documents
STATISTICA Graph documents contain all options, features, styles, and information about the inserted, linked, or embedded objects, as well as all the relevant data, and therefore, they can be shared between users even if they are separated from their original datasets. Moreover, you can reopen a graph document at a different time and on a different computer, and continue to analyze or customize the graph, e.g., change fitting options, categorization settings, etc.
Dynamic Links Between Graphs and the Input Data
All graphs created from the spreadsheet data can automatically maintain their links to the data. For applications in exploratory data analysis, a macro system can be defined to have a series of predefined graphs automatically recreated for each of a series of datasets; all graphs can be automatically printed, saved, or directed to a presentation quality report combining graphs with text.
Graphs Automatically Updated in RealTime
If the spreadsheet is linked to an outside data source, graphs can be set to update automatically whenever the spreadsheet links are updated (e.g., from a remote data warehouse or set different databases as defined using STATISTICA Query). For example, quality control graphs (see the illustration) or other types of graphs can be used for realtime monitoring of specific quality indices or to control the progress of a laboratory process. STATISTICA is compatible with virtually every data acquisition system, and the data can be transferred to STATISTICA either via a variety of links to the spreadsheet (that can be updated in the background) or macro emulated keyboard input. If the spreadsheet contains formulas to transform or “clean” the input data, the respective parts of the spreadsheet can be set to automatically recalculate whenever new data are received, and then the transformed data will be sent to the chart. Also, those types of systems of practically unlimited complexity can be custom defined in STATISTICA Visual Basic.
Selection of Graphs
Optimized access to graph type selections.
The list of all types of graphs supported is long; however, access to all types of graphs in STATISTICA is designed to minimize potential confusion among the number of choices, and to guide you in the process of selecting graph types. All choices can be made from menus, toolbars, or a convenient Graphs Gallery facility (designed to simplify access to the many types of graphs, and integrated with the STATISTICA Help) custom macros can also be used. See also the next topic on shortcuts.
Quick access to common types of graphs, shortcuts.
The most commonly used types of graphs are available via quick “singleclick” facilities [e.g., Quick Stats Graphs (see the illustration at left) or basic selections of Block Graphs] accessible from shortcut menus or the top sections of other menus, and are designed to reduce the number of necessary selections to the very minimum. For example, these quick access facilities will skip all option dialogs, apply dynamically determined system defaults for all settings, and if a block of values is highlighted in the current spreadsheet, will even suggest the selection of variables from the block. Other alternative ways of accessing general types of graphs are also supported; for example, you can assign the most commonly used types of customized (or customdesigned) graphs to buttons on existing or custom defined toolbars or shortcut keys, or append them to a menu.
A review of main types of graphs.
The selection of types and subtypes of graphs that can be produced by STATISTICA exceeds the limits of what would be reasonable to describe in detail in this overview. This section includes only a summary of the main types of graphs.
In addition to the large selection of styles of “business type” (Microsoft PowerPoint or Excelstyle) 2 and 3dimensional graphs (such as bar, column, pie, line, area, stacked, stock market style HighLowClose, range, deviation, block, etc.), STATISTICA offers a wide variety of statistical (analytic and exploratory) graphs. These include histograms (including smoothed, cumulative, multiple, doubleY, fitted, hanging histobars, one and twoway categorized multihistograms, matrix histograms, 3D bivariate distribution histograms and surface smoothed histograms, histograms coordinated with scatterplots, histograms coordinated with sequence, moving average, etc., graphs, and many others), means plots with error bars, box plots, range plots, boxandwhisker plots [including userdefined plots of central tendencies (such as trimmed means or medians) and measures of dispersion (userspecified percentiles, ranges, SDs, SEs, or nonoutlier ranges], plots of outliers and extreme values, box plots with outliers coordinated with scatterplots, as well as 2D, 3D, and categorized plots of ranges or variation in multiple series or groups of data, and many others], scatterplots [including multiple, doubleY, frequency, marked, pointlabeled, weighted, Voronoi, one and twoway categorized multiscatterplots, scatterplot matrices, XYZ, ternary, Polar, multiplesubset where each subset of data is specified by customdefined selection conditions, compound scatterplots with coordinated histograms or box and whisker plots, scatterplots of quantiles for comparing distributions, and many others, including such specialized types as scatterplots where every data point is represented by a custom graphics file (BMP, WMF, JPG, PNG) or iconscatterplots where data points are marked by multidimensional icons visualizing relations between additional variables, e.g., small pie charts or Chernoff faces], line plots (including multiple, aggregated, marked, case/profile plots with range lines, doubleY, 3D, ribbon, and many others), a wide variety of 2D and 3D function fitting and function plotting graphs, and distribution fitting graphs (including flexible implementations of categorized and regular QuantileQuantile and ProbabilityProbability plots for a wide variety of distributions); categorized and regular Missing/Out of range graphs mapping distributions of missing data (or values from userdefined ranges) across datasets or their subsets defined by categorizations; icon plots (including flexible implementations of Chernoff faces, star, sun ray, polygon, pie, column, line, and profile icons), matrix plots (including square and rectangular scatter, line, and column matrices, and multiplesubset matrices), 2D and 3D triangularcoordinate (ternary) graphs [including scatterplots, contours (projected and 2D), surfaces (with a selection of ternary fitting options), ternary space and deviation graphs, as well as categorized (multiple) 2D and 3D ternary plotdisplays of all supported types]; one and twoway categorized multiplots [that enable you to generate flexible “crosstabulations” of multiple scatterplots, histograms, line graphs, function fitting graphs, boxandwhisker graphs, pie charts, various bar, column, range, deviation, tessellation and other specialized plots and their combinations, including a wide selection of multiple (i.e., “categorized by…”) 3D graphs]; a wide selection of 3D graphs, including various XYZ scatterplots, space plots, spectral plots, deviation plots, ribbon plots, block plots, box plots, 3D range plots and multiple range plots (including flying boxes, flying blocks and broken blocks, flying doubleribbon ranges, flying border, point, and error bar ranges); various userdefined surface plots with projected contours, line plots, trace plots, contour plots, function plots, 3D pie charts, and others; special multiple3D (4D) graphs, and many others. 3D displays feature onscreen rotation, control over all proportions of the respective solids, perspective, etc. This also includes the 4D graphs (e.g., multiple surface plots can be rotated simultaneously on the screen in usercontrolled perspective).
UserDefined Graphs
All graph definition and customization facilities in STATISTICA can be used to define new types of graphs, which can be permanently added to the floating toolbars or appended to the menu. There are several ways in which new graphs can be defined.
The simplest method to customdefine a new graph (to be used repeatedly) is to create it as usual, using any options from the respective graph definition dialogs, and then click the Add as a Userdefined Graph to Menu button. STATISTICA will prompt you for the name (and optionally, if the selection of variables is to be preserved), and the new graph name will be added into the menu.
An alternative method to customdefine a new graph (for repeated use) is to create a macro (either by recording it or writing a program using the STATISTICA Visual Basic editor). The new macro can then be assigned to a button on a toolbar, menu option, and/or a shortcut key.
Finally, STATISTICA Visual Basic offers comprehensive access to all graphics procedures of STATISTICA. Creating even complex graphs with STATISTICA Visual Basic is surprisingly easy. For example, you can start by recording a macro by creating a graph interactively and then customize the code. A Function Browser is provided to speed up entry of the graphics functions. Graphs of any degree of complexity can be custom defined in STATISTICA Visual Basic and assigned to buttons, menu options, and/or shortcut keys for repeated use. The applications are virtually countless and range from simple line graphs (e.g., created for every case in a data file and overlaid in one display) to, complex types of specialized graphs, technical drawings, and diagrams related to datasets. Completely new graphics structures can be created using the drawing tools (graphics primitives). STATISTICA Visual 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 menudriven graph customization procedures). STATISTICA Visual Basic can create complex compound graphics documents with ActiveX/OLE links (including nested documents), diagrams related to data (that can be updated/redrawn by pressing a button), and many other types of displays. The graphics output can be directed to all output channelsstand alone windows, workbook, and/or reports; they can be automatically printed, saved, or combined with output from other applications (e.g., output into a Word document or an Excel spreadsheet.) Your new, customdesigned graphics procedures can be permanently added to the STATISTICA system by assigning them to any controls. Simple to use, predefined dialogs can be set up (with the intuitive, interactive dialog painter) to produce customized “front ends” for these new procedures (e.g., prompting the user to enter the necessary parameters or text of titles, select specific options, or any other type of input).
Flexibility of Graphical Representations of Data: An Example
One of the unique features of STATISTICA is its facilities to flexibly experiment with different graphical representations of the same set and arrangement of data. After the selected graphical representation of a dataset is displayed on the screen (e.g., a 3D scatterplot), the type of graphical representation and the layout style can be flexibly adjusted to achieve the desired analytic or presentation effects.
For example, after a 3D scatterplot is produced, you can interactively convert it into a space plot (showing deviations of individual data points from a fixed plane), “compress” the data points into a userrequested number of spectral planes, fit a surface to the data points (choosing from a variety of surface types and display styles, stiffness of the fit, etc.), compare the fit of data to surfaces produced from customdefined functions, experiment with refitting the surface after interactively removing and restoring outliers (with a 3D slicer or a cube brush), perform onscreen rotation and adjust the perspective of the graph, zoom in on specific concentrations of data (e.g., scroll in logical zoom mode to create an effect of moving a “strong magnifying glass” over the graph to explore specific areas), interactively identify specific points by labeling them with one of the many brushing methods supported, etc.
Graphical Data Analysis
STATISTICA features a comprehensive selection of tools for graphical data exploration and analysis, and an extensive set of facilities to identify relations, trends, and biases “hidden” in unstructured datasets. The analytic techniques include function fitting and plotting, data smoothing, overlaying and merging of multiple displays, interactive categorizing of data, splitting/merging subsets of data in graphs, aggregating data in graphs, identifying and marking subsets of data that meet specific conditions, shading, plotting confidence intervals and confidence areas (ellipses), generating tessellations, layered compressions, spectral planes, and projected contours, data image reduction techniques, interactive (and continuous) rotation of 3D graphs, selective highlighting of specific series and blocks of data, a uniquely powerful and comprehensive selection of brushing techniques including a flexible “animated brushing” facility (see the Brushing topic, below), 3D slicers and interactive, fully customizable cube brushes), analytic zooming tools that enable you to interactively create subgraphs by selecting an area (or a cube) on an existing display, and many others.
Point markers on plots can be made transparent with an interactive transparency control with onscreen sliders (requires Windows Vista SP 2 or Windows 7). Transparency control is a powerful graphical exploratory technique that enables you to reveal trends hidden in the dense concentrations of data points (especially scatterplots and scatterplot matrices generated from extremely large data sets).
Fitting, Smoothing, Overlaying
Specialized smoothing and fitting methods related to particular statistical procedures are available as part of output selections in the respective statistics modules. However, a comprehensive selection of generalpurpose smoothing and function fitting methods are available at any point of your analysis as part of the general graphics options, and they include a wide variety of distribution fitting options (including Beta, Exponential, Extreme Value, Gamma, Laplace, Lognormal, Lowess, Normal, Poisson, Rayleigh, and Weibull), as well as standard fitting and smoothing procedures including linear, exponential, logarithmic, spline, normal, polynomial (of userselectable order), bicubic spline, distanceweighted least squares smoothing, negative exponentiallyweighted smoothing, ternary linear and quadratic, ternary cubic and special cubic. Userdefined 2 and 3dimensional functions (as well as sets of parametric curves e.g., to draw a circle or an ellipse) can also be plotted and overlaid on the graphs. The functions can reference a wide variety of distributions including Beta, Binomial, Cauchy, Chisquare, Exponential, Extreme Value, F, Gamma, Geometric, Laplace, Logistic, Lognormal, Normal, Pareto, Poisson, Rectangular, Rayleigh, Student’s t, and Weibull, as well as their integrals and inverses.
Fitting Arbitrary Functions
Additional facilities to fit userdefined functions of practically unlimited complexity to the data are described in the section on Nonlinear Estimation. The function minimization can be performed using a selection of powerful fitting algorithms (including LevenbergMarquardt, quasiNewton, Simplex, HookeJeeves pattern moves, and Rosenbrock pattern search method of rotating coordinates), and according to the default or userdefined loss functions.
Brushing Techniques
Overview
The comprehensive implementation of brushing techniques (for exploratory data analysis and hypothesis testing) includes a wide variety of data selection and identification methods as well as various options to manage the selected data. Due to STATISTICA’s proprietary graphics technology (see the topic on technology, below), the brushing facilities are extremely responsive even on large scatterplot matrices showing large datasets. Point, rectangle, lasso, slice, and cube brushes can be used; brushing is supported for all categories of graphs (including 2D, 3D, categorized, scatterplot matrices, and even such specialized displays as graphs with Polar coordinates or categorized 3D scatterplots with triangular coordinate systems). Brushing actions supported include interactive labeling, marking, elimination/suppression, reversing all operations, and changing the status of individual or globally selected sets of data points (e.g., depending on combinations of conditions met by the data points).
Animated Brushing
Animated movement of multipoint (area, slice, or cube) brushes (especially useful in exploration of scatterplot matrix displays) enables the user to watch the dynamics of relations between variables in multivariate datasets. For example, a rectangle brush covering 10% of the range of the variable INCOME can be set to “flow” over the entire range of INCOME (at a speed interactively controlled by the user). As the brush “flows” over the ranges of INCOME, all data points that belong to the currently covered (“brushed”) ranges of INCOME will automatically “light up” (i.e., become highlighted) in all scatterplots of the matrix, allowing you, for example, to inspect the contribution of observations representing specific levels of INCOME to the relations between all other variables in the dataset.
Rangebased brushing
In addition to the mousebased methods, flexible tools are available to interactively brush/select subsets of data by specifying ranges of values and/or combinations of attributes of data points. Also, facilities are provided to manage the selected data points (e.g., selectively copy them to the Clipboard, copy or move them to a new column/plot, etc.).
The Technology Supporting the Graphics Procedures, Speed (A Technical Note)
The unique technology behind all graphics procedures in STATISTICA not only enables advanced exploration and visualization of data and facilitates the analyses listed in the previous topics, but it also contributes to the overall responsiveness of the program. For example:
 All graph redraws are performed at speeds limited only by the hardware of the computer and the drivers (see the sections on Speed and Comparative Benchmarks).
 STATISTICA uses the multithreading technology for all graph redrawing operations (thus, complex displays are redrawn in the background while you continue with the next operation). Custom resolution enhancement procedures are used to minimize graph distortion, optimize the graph appearance, and increase the readability of fonts and markers and the accuracy of the display.
 In addition, dataimage compression and/or display density filters can be used to increase the readability of graphs based on large datasets.
 A “logical” zoom (i.e., a zoom revealing new details and not merely stretching the display) can be performed on all relevant graphic displays and options are provided to create subgraphs from interactively selected 2D or 3D subsets (or slices) of data.
 The implementation of brushing in STATISTICA is based on proprietary graphics and image processing technology (which is largely responsible for the overall speed of screen redraws and responsiveness of the program). Even with moderate to large datasets, all operations are performed with virtually no noticeable delay. The proprietary region selection and virtual image technology enables you to perform interactive brushing even with extremely large datasets, on large screens (e.g., 1600 x 1200) with unlimited numbers of colors, and the brushing can be performed in real zoom. The proprietary region selection procedures supporting regions of unlimited size enables these procedures to break the system region size limitation even on the older versions of Windows.
 When the zoomout or reducing the window size operations make the fonts on the graph too small to read, a unique toolbar tool is provided to globally and proportionately adjust the display sizes of all graph fonts and other components set in an independent metric (e.g., point marker size) for increased readability or presentation effects (see the illustration at left). The same options are also supported for compound documents.
 Crosssections of 3D graphs can be interactively reviewed “slice by slice” with a unique brushing slicer.
 By default, all graph windows are resized while automatically maintaining the graph aspect ratio (as in CAD programs). The user has full control over the graph mapping options, including the mapping base, proportions, margins, graph aspect ratio, translation of logical sizes of graph components as set in points (one quarter of a point is 1/288 of an inch) into physical sizes as they appear on displays, printouts, or in graph regions when the graphs are created in the STATISTICA Report editor or pasted into documents in other applications.
 The ActiveX/OLE support includes simultaneous support for server and client mode; compound embedded objects (up to fourth level) are supported (see the illustration at left); proprietary technology is used to protect against circular crossreferences between linked objects; the server mode offers interactive control over graph mapping of inserted objects.
ActiveX/OLE
Overview
STATISTICA offers a comprehensive implementation of both client and server ActiveX/OLE functionality. The server facilities offered in STATISTICA enable STATISTICA graphics objects to be embedded or linked into other applications. For example, a STATISTICA graph pasted or pastelinked into a word processor can be easily edited (using all STATISTICA customization facilities) by doubleclicking on the graph in your word processor document. Those OLE server facilities are supplemented by comprehensive support for OLE client services (see the below). All STATISTICA graphics documents can serve as clients (ActiveX containers) for objects from other applications (or STATISTICA’s own objects). Foreign files can be inserted and opened directly into STATISTICA Graphs (and optionally linked to their sources to enable automatic updates when the source files change). ActiveX/OLEcompatible applications can also be called directly from within STATISTICA graphics documents to create new objects to be inserted [e.g., equation editors (as shown on the illustration on the left) and other specialized applications or word processors, spreadsheets, etc.]. The nesting of objects can be up to 4 levels deep (i.e., an object can be embedded in an object, that is embedded in another object, that is embedded in another object). For example, you can paste into STATISTICA graphs foreign objects that are containers for other containers. All ActiveX/OLE objects in STATISTICA documents can be linked, and they can be updated dynamically; proprietary technology is used to protect against circular crossreferences between STATISTICA OLE objects even on older Windows systems. Both ActiveX/OLEcompliant and incompatible fileobjects can be dragged (across application windows) directly from the Windows Explorer and dropped onto STATISTICA Graph windows for further editing and adjustment (as shown in the illustration on the left).
Applications
The ActiveX/OLE functionality saves time and offers obvious advantages when STATISTICA Graphs are used in other applications (e.g., word processor documents).The client options offer powerful tools to build customdesigned artwork or dynamically updatable and/or efficiently “compressed” documents (e.g., such as compound documents mentioned in the previous paragraph). Also, many new types of STATISTICA graphs can be built using this technology. The compound graph document facilities provide the user with great flexibility for building new types of combination graphs and graphical displays. They also offer the most straightforward and intuitive access to all graph customization facilities and methods to update the documents; updating the component graphs is as easy as clicking the mouse. Compound documents with embedded or linked objects can also be created in STATISTICA Visual Basic.
Datasets in Graphs
STATISTICA can produce graphs from datasets of unlimited size (e.g., scatterplots with hundreds of thousands of data points can be created to explore patterns of outliers). In order to help create readable graphs from extremely large datasets, an option is provided to use a fast dataimage reduction procedure to generate representative “compressions” of large numbers of observations. STATISTICA, by default, takes advantage of all relevant information in the data file, including all alphanumeric (text) values that can be used for labeling graphs (categories, subsets, individual observations, etc.), including their incell formats and special attributes. Also, facilities are provided to use values from any variable (or case names) as labels for data points, categories, or subsets in all graphs. When graphs are linked to their respective datasets, not only numeric data values, but also text values, labels, titles, etc., derived from the original dataset can be automatically updated when the source dataset changes.
Graph Customization
Practically every detail of the appearance of the graph (hundreds of specifications) can be controlled by the user and accessed directly from the screen (see the following topics on Mouse support), including the location, style and length of minor tick marks to the overall proportions of the graph, and its position on the page. The adjustments can be performed with a minimum number of mouse clicks or keystrokes (see the section on Shortcuts), and they also can be converted into permanent defaults for the particular type of graph (default “style sheets”); libraries of such graph styles can be maintained (see the section on Configuration). Comprehensive facilities to control the sizes and patterns of all graph components are provided. The line width and marker sizes can be adjusted in fractions of a point (1/4 of a point is 1/288 of an inch). These size control facilities can be globally and interactively adjusted by controlling the graph mapping options (that is, the manner in which the metric of user settings for sizes of objects gets transformed into physical sizes of the respective objects in the display or printout). In addition to large selections of predefined patterns (e.g., 32 predefined fill patterns), unique facilities are provided to custom design twocolor styles of linepatterns, point marker patterns, and fill patterns (e.g., a navy blue circleshaped point marker with a gray inside fill pattern or a dotted line consisting of red and blue dots). Specialized predefined (default) color palettes are provided for particular graph components. Moreover, flexible facilities are provided to custom design new palettes using unlimited numbers of colors as supported by the current device.
Scales
A large number of options enable the user to control every aspect of the scales. For example, support is provided for multiple (parallel) scales. Scales can feature multiple (true) scale breaks that can be used to “compress” specified segments of the display. Scale values can be placed at arbitrary locations, and their format can be controlled using a selection of options. Moreover, facilities are provided to automate tedious aspects of scale definitions; for example, date scale values can be created automatically, STATISTICA can be instructed to display only every n’th scale value, and specific aspects of the scale definitions can be automatically transferred to the opposite scale or to all scales. Note that all those graph customization, drawing, object management, etc. facilities are fully supported in STATISTICA Visual Basic.
Interactive Scaling
You can directly interact with the scaling on the graph by hovering the mouse pointer about the axis labels toward the end of the axis and pulling left or right to change the scaling. Interactive scaling in a powerful graphical exploratory technique that enables you to reveal hidden trends by stretching or compressing the desired parts of the display. Similarly, Interactive panning allows you to pan to the right or left by hovering the mouse pointer above the axis labels toward the center of the axis.
►
The Comprehensive System for Managing Graph Styles
One of the general strengths of the graph customization facilities in STATISTICA are the options to automate and speed up repeated custom operations. Practically all types of customizations can be saved in form of custom styles (with arbitrary names), and any of these styles can also serve as global or local defaults; this applies to hundreds of specifications. For example, even each of the drawing tools can be customized to avoid repeated adjustments (e.g., sizes, colors, patterns, scaling, locking); complete definitions of individual scales (e.g., with custom positioned scale values with special subscript formatting and bold prefixes, patterns, formats, breaks, etc.) or complete sets of all scales (for a graph) can be saved as reusable templates or copied and pasted via the Clipboard as custom objects. Custom definitions of surface or contour levels (with specific value ranges, patterns, colors, palettes, etc.) can be saved as reusable templates. More global sets of specifications (new, partial, or complete graph styles) can be saved as menu or toolbar options (see the section on Shortcuts). Also, custom selections of specific graph customizations can also be converted into “singleclick shortcuts” by storing them as macros (either recorded or edited) and assigning them to buttons on local or global toolbars. Moreover, custom types of graph customizations can also be defined (and assigned to buttons) using STATISTICA Visual Basic.
Mouse Support
Comprehensive Left mousebutton support
All graph customization options can be accessed directly from the display by clicking on the respective graph components or objects. Unlike rightclicking (which displays a menu of all relevant customization dialogs for the object, see the next topic), a doubleclick with the Left mouse button directly brings up the most commonly used customization dialog for the object (see the illustration). A singleclick with the Left mouse button will select (i.e., highlight) a graphic object to be modified (e.g., dragged, resized, rotated, see below). A singleclick on a data point will select (highlight) all data points that belong to the current series, such as subsets of data in multiple or multiplesubset scatterplots (see the topic on Graphics technology). Other operations that can be performed with the Left mouse button include drawing, embedding, unlocking objects, zooming, various types of brushing, scrolling graph magnifications, rotating text, etc. The mouse can also be used in all graphs to select the respective data series (or display the local Graph Data Editor containing the data values; this option is supported for all graphs, even those that display derived or computed data, e.g., probability plots).
Comprehensive Right mousebutton support
As everywhere else in STATISTICA (e.g., spreadsheets), rightclicking on an object will display a shortcut menu of all available categories of operations that can be performed on the selected graph component or object (see the illustration). This facility allows you to avoid going through hierarchies (multiple levels) of dialogs by making the complete list of all lastorder dialogs available directly. If you rightclick outside any specific objects or graph components, a global shortcut menu for the graph will be displayed when you can select from among several global options.
Smoothness and precision of dragging, brushing, drawing, etc.
As mentioned before, the mouse is used in STATISTICA to perform a variety of interactive operations on graphic objects (such as dragging, proportional and nonproportional resizing, realigning, rotating, stretching, selecting/highlighting, zooming, drawing, etc.). The technology supporting all of these operations in STATISTICA ensures “smooth” and precise graphcustomizations and adjustments. All of them can also be performed in zoom mode (see below) to further increase the precision. Custom resolutionenhancement redrawing procedures are used to deliver clean and precise outcomes regardless of the size and proportions of the window, the degree of magnification (zoom), or the graph mapping mode. Also, keyboard keys can be used to emulate the interactive mouse operations on highlighted objects, and when the Ctrl key is pressed, the movement can be performed in the smallest increments supported on the current device (one pixel).
UserInterface Shortcuts
Graph creation
The number of mouse clicks normally required to produce even elaborate and customized graphs is reduced to the very minimum. However, the necessary user input can be reduced and simplified even further by:
 defining new, complete graph types and adding them permanently to the menu or assigning them to buttons on toolbars or to shortcut keys (both options are illustrated in the screen at left);
 using the internal batch processing facility (which enables you to create and automatically print, save, or insert into a report, long sequences of graphs, e.g., the same type of graph for each variable in a long list of variables);
 using recorded and/or edited macros to create specific graphs (which can then be assigned to buttons on the global or local toolbars or to shortcut keys);
 specifying graphs via the STATISTICA Visual Basic language; this can include even long sequences of highly customized graphs, compound graphs, custom drawings, diagrams, or other displays connected to data or interactively controlled by userdefined input; these custom graphics procedures can also be assigned to buttons on the global or local floating toolbars (see the illustration at left) or to shortcut keys;
Graph customization
The appearance of even the default graphs in STATISTICA is very refined and, thus, the graphs are defined by hundreds of settings. The layout and appearance of each of these graph components can be independently customized. The number of mouse clicks normally required to produce even elaborate customizations of graphs is reduced to the very minimum. However, the necessary user input can be simplified even further by:
 using predefined toolbar buttons offering direct (“singleclick”) access to a variety of global operations, such as zoom, plot region, margins, proportions, mapping mode, or simultaneous proportional resizing of all fonts, markers, etc., in the current graph and/or all fonts, markers, etc., in embedded graphs;
 using the predefined system shortcut keys and other keyboard shortcuts;
 using the userdefined system of styles (mentioned above) to adjust the permanent global or local defaults; libraries of such custom styles and default styles can be maintained;
 merging graphs, which offers another way to transfer all customizations from one graph to another;
 creating reusable objects/templates/styles from specific parts of a respective graph (e.g., all scale definitions including customformatted scale values, etc.; all definitions of levels and shading in a surface plot or contour plot);
 defining new, customized graph types and adding them permanently to the menu or assigning them to buttons on the global or local floating toolbars (see illustration at left) or shortcut keys;
 using recorded and/or edited macros to define specific types of graph customizations (which can then be assigned to buttons on the global or local floating toolbars or to shortcut keys);
 specifying new graph customizations in the STATISTICA Visual Basic language; this can include customizations of compound graphs, custom drawings, or diagrams; the customizations can be interactively controlled by userdefined input entered into dialog boxes defined with the STATISTICA Visual Basic interactive dialog editor; these customdefined graph customizations (e.g., placing a text label “xyz” rotated by 25° at a specified position in the graph) can be assigned to buttons on the global or local toolbars (or to shortcut keys) to create sets of customdefined customizations.
Creating multigraphics displays (AutoLayout Wizard)
An often tedious task when creating complex multigraphics displays is to precisely arrange multiple graphs or objects on one page or slide, and to adjust their scaling. One of STATISTICA’s Wizard facilities (the AutoLayout Wizard, see the illustration) was designed to simplify this task and to accomplish it for you with a minimum number of mouse actions. You can select saved or unsaved graphs from all currently open STATISTICA modules, or files saved to the disk. Then pick up a layout from a gallery of preselected layouts (suggested by STATISTICA depending on the number of objects you have selected, see the illustration at left); at this point, you can take advantage of facilities to adjust scaling, margins, auto titles, footnotes, etc. Now, when you click OK, STATISTICA will create perfectly aligned and proportioned artwork. Custom “menus” of layouts created with the help of this Wizard can be placed on toolbars for repeated use (via “singleclick” access).
Graphic Text Editor
Practically unlimited amounts of text (including long reports with embedded graphs) can be inserted into STATISTICA graphs as active ActiveX/OLE objects, metafiles, or as text. An integrated, WYSIWYG graphic text editor can be used to create even highly customized text and labels. The editor includes its own specialized toolbars and offers not only the standard text formatting features (complete font control, text justification, borders, background, etc.) but also options to insert symbols from any character set (e.g., Greek letters). Options are provided to insert autoupdatable fields including text of specific equations fitted to the data, legend symbols for specific plots within the graph, levels for contour lines and areas (shading), etc.; this facility enables you to set up custom autoupdatable legends and lists of functions (e.g., for categorized graphs or graphs with multiple plots). The anchor point can be adjusted independently of the text justification (e.g., text centered within its box can be anchored by its lowerright corner) and tools are provided to rotate the text in 1o increments, either by specifying the angle or by rotating the text interactively on the screen (when you are rotating the text interactively by dragging its corner on the screen, the current rotation angle is displayed in the toolbar show field).
Configuration, User Preferences
As mentioned before, by using the flexible system of styles, practically all graph customization settings can become permanent program defaults (global or local, for a specific project) that will affect new graphs. This applies not only to sizes, colors, patterns, styles, backgrounds, scales, fonts, titles, etc. of almost countless specific graph components, but also to such global features as the way in which graphs are scaled and mapped into the plot region, the margins within graph windows, window proportions, etc. As mentioned before, in order to facilitate the setup of desired configurations of graph defaults, options are provided to automatically “copy” all the settings from the current active graph into a style, and then they can be modified further or saved as selectable templates into your library of alternative graph styles.
Drawing, OnScreen Customization
Unlimited numbers of objects can be added to every graph via an extensive selection of linking and embedding facilities and a comprehensive set of onscreen drawing tools is available at any point while viewing graphs. The tools include elementary objects such as lines, rectangles, polygons, ovals, rounded rectangles, arcs, etc., as well as specialized facilities such as freehand drawing tools producing editable objects, userdefined styles of arrows (of practically unlimited shapes and styles), a userdefined “error bar” tool, and a flexible WYSIWYG text editor supporting multiline text and special formatting options (see the Graphic text editor topic above). All drawn or imported graphic objects remain “active” and modifiable and you can always customize them in a variety of ways. All objects can be interactively resized, repositioned, or dynamically related to particular graph locations. Also, you can change colors of objects, their line patterns, fill patterns, line widths, and backgrounds, place frames around them, attach labels, etc. All custom editing, pattern, and size selection facilities offered by STATISTICA (see the Graph customization topic, above) can be applied to STATISTICA’s custom graphic objects. Many special, CADstyle drawing and object manipulation facilities are provided for precise analytic or presentation effects. For example, STATISTICA graphic objects can be proportionately resized or, stretched in one direction, or their shapes can be customedited by adjusting individual microsegments (curvecomponents) of the already finished freehand drawings (e.g., if a freehand drawing has the shape of an apple, you can “pullin” its sides to make it look more like a pear). The microadjustments can be performed in zoom mode (see below). The Clipboard can be used to facilitate drawing and to build compound objects; special internal Clipboard representations are supported to copy complex object structures in a way that is transparent for the user. One click of the mouse on a toolbar button will allow you to adjust the default “redraw order” for every custom graphic object (e.g., to pull a hidden object to the front); this also allows you to achieve special presentation effects (e.g., you can experiment with adding different full or partial backgrounds to an existing graph) or to meet special analytic goals (e.g., you can insert a component of one graph “underneath” another for graph comparisons). The selection of these and other facilities of the STATISTICA drawing system (see below) was designed to make sure that you will never need a “designated presentation graphics” or “diagramdrawing” program.
Drawing in Zoom Mode, Special Mouse Actions, Alignment of Objects, Customized “SnaptoGrid”
Drawing and manipulating objects on the screen and changing their styles and attributes is as easy as moving the mouse. The shortcut menus associated with objects (and invoked by rightclicking on the object) speed up the access to specific customization facilities. At any point of drawing or creating technical diagrams, you can switch to the zoom mode and draw “under a magnifying glass” to achieve superior precision and gain access to small details. You can also switch to the scroll graph area in zoom mode and effectively examine the graph under a magnifying glass (or perhaps we should say, a microscope). You can interactively stretch not only the entire graph (while optionally maintaining its proportions), but also selectively add space in any of the margins (e.g., for embedded objects, comments, etc.) by using another area adjustment tool (accessible by clicking a button on the toolbar). Facilities are provided to precisely align objects (in fixed window coordinates or dynamic graph scaling units); all graphic objects can be attached to specific graph or window coordinates. Also, a userdefined, customizable “snaptogrid” facility is provided and can be enabled/disabled via shortcut keys (the display of the actual alignment grid can also be toggled by pressing a shortcut key).
Quality of Graphical Displays, Resolution, Fonts
STATISTICA offers the highest quality and precision of graphical output supported by the currently available hardware. In fact, the program internally generates all graphical displays at a higher resolution than what is available in existing output devices. All scaleable fonts and symbol sets are supported in 3dimensional displays, the fonts are transformed in the respective planes of the 3dimensional space and according to the usercontrolled 3D perspective (nontransformed fonts can also be selected); proprietary technology is used to achieve the highest quality of 3D transformations of fonts (see the illustration). Also Postscript fonts can be used in graphs and transformed in 3D perspective. Every predefined category of graph features a separately designed set of sizecoordinated and dynamically adjusted labels, titles, and legends, all of which can be easily resized and edited by the user.
Saving, Exporting Graphs
All graphs can be saved into active STATISTICA graphics documents (ActiveX objects) that can be placed on the web, embedded in other applications as active inplace editable objects, etc. They contain not only all the graphics components of the objects (including all their customizations, drawings, embedded objects, references to linked external files, links to raw datasets, etc.), but also all the respective datasets, thus allowing the user to continue interactive graphical data analysis (brushing, fitting, smoothing, rotating, editing of data, changing categorizations, selected subsets, etc., and all onscreen objects). All graphs can also be saved into a variety of graphics file formats allowing you to exchange artwork with other graphics applications without using the Clipboard. Supported formats include JPG, PNG (a new version of GIF), Windows metafiles (WMF), deviceindependent bitmaps (BMP), Postscript files (EPS [print to]), and others.