Getting Started with TIBCO® Data Science - Workbench (formerly Statistica)

Last updated:
8:23am Sep 09, 2020

The following e-learning materials are available free of charge. They provide helpful hints and guidance to navigate the software more efficiently. You may see Statistica referenced as TIBCO Data Science Workbench or Statistica within videos and documentation. This is the same application. 

IMPORTANT: There are two user interfaces within the Statistica application. 1) Interactive; data is visualized within a spreadsheet, analysis is completed on the spreadsheet by selecting menu items. 2) Workflow; no/low code sequence of steps (retrieve data > transform data >  analyze > write results to database, etc..) to complete an analytic project, workflow is visualized within a workspace object and can be scheduled to execute, run ad-hoc within Statistica or execute within the Spotfire user interface. 

Trial

Are you actively trialing Statistica? Check out the TIBCO Statistica Trial Toolkit for all the tools you'll need to ensure a successful trial and enjoyable trial experience!

Videos

Looking for videos to get started?

Some helpful videos (with some walk-throughs) include:

Documentation

Electronic Statistics Textbook is recommended by Encyclopedia Britannica for its "Quality, Accuracy, Presentation, and Usability."This on-line textbook covers Statistical theory and how to interpret results.

Product Help Examples - Additional examples and information can be found in the help documents that come with your software. After starting the Statistica application, type F1. Look in the Contents tab. Read the Feature Finder article, then explore the over 70 step-by-step examples for Statistica and Statistica Enterprise Manager. You can also search, using the keyword Example, to locate these articles within the help.

Statistica Object Model Documentation - This help document is attached to bottom of this page; sdn.zip. It is also installed your software. After starting the Statistica application, type F1. Look at the bottom of Contents tab for Statistica Object Model. Virtually every detail of data importing, visualization, cleaning, data analysis, predictive analytics, and reporting can be controlled via a scripting language. Macros can be recorded within Statistica and saved for re-running later. Business logic can be stored within Statistica Enterprise Manager. These objects can be called directly from many other applications or COM-compliant programming languages (e.g., C#, C++, Java, VB.NET).

Examples & Templates

Examples can be downloaded from TIBCO Exchange

Statistica also ships with workspace examples & templates to help get started on a project. Examples contain data that connects with a workflow to complete various analytics; data cleaning to modeling building. Templates do not include example data.

To open an example start Statistica. Select Home menu ► pull down arrow on Open menu ► Open Examples menu. Browser to the Workspaces folder.

Open Basic_DM_Example to see a simple example. Select the Run All button to execute the workflow and explore the different steps. This executes model building for three different algorithms. The training and testing workflows will decide which model is the best. Delete the CreditScoring dataset and connect to your data.

Data Cleaning, Transformation, Feature Selection

Here is a short list of data related examples.

  • Anonymize Variable Names
  • Automated Data Cleaning
  • Create Design Matrix
  • Feature Selection with Spark
  • Feature Selection with Statistica
  • Lasso Regression
  • Spotfire Data Science Import; call Spotfire Data Science workflow and return results to Statistica for metadata analysis

Champion Challenger

Here are the examples that build multiple models and select the best performing model.

  • Advanced Comprehensive Classifiers; General Discriminant Analysis (GDA) Models vs C&RT Standard Trees vs Standard CHAID vs Exhaustive CHAID vs Neural Network
  • Advanced Comprehensive Regression Models; General Regression Models (GRM) vs C&RT Standard Trees vs Standard CHAID vs Exhaustive CHAID vs Neural Network
  • Customer Churn Analysis; C&RT Standard Trees vs Boosted Trees vs Generalized Linear/Nonlinear Models (GLZ)
  • General Forecaster Neural Network Time Series; Simple Exponential Smoothing and Forecasting vs Exponential Smoothing and Forecasting vs ARIMA Models and Forecasting vs Neural Network
  • Spark Model Comparison; Statistica Random Forest vs Spark Random Forest
  • Spark Regresion; Linear Regression vs Logistic Regression vs Generalized Linear Model (GLM)
  • Spark Trees; Decision Tree vs Random Forest

Horizontal or Vertical Solutions

  • Credit Scoring WoE
  • Customer Churn Analysis
  • Consumer Preferences
  • MNIST Digit Recognition with Microsoft CNTK
  • MNIST Digit Recognition with H2O Deep Learning
  • Non Linear Time Series for R - integration with R
  • Optimization - Bayesian Optimization for Binary Response
  • Optimization - Bayesian Optimization for Continuous Response
  • Transportation / logistic delay prediction with H2O Deep Learning
  • Website visitors association

Other

  • H2O DRF Distributed Random Forest
  • H2O K-Means
  • H2O PCA Principal Component Analysis
  • General Classification Quick Linear Models
  • General Quick Linear Models
  • SPC Charts With Alarming
  • SPC Charts

Roles Based Security

Other Resources

Are you an existing user? Try to access TIBCO Support portal where you can find additional useful information. From there, you can reach Knowledge base articles. Apart from solutions to frequent asked questions, you might find useful articles showing you tips and tricks with the software functionalities such as: How to embed document in Workspace, How to access help for specific dialog, How to connect to Aster database, What are the cases and variables, etc.

If you are not able to find information you are looking for, please use Answers Section, insert your questions and tag product as TIBCO Statistica.

Back to main Statistica wiki page

Attachments

AttachmentSize
Package icon sdn.zip4.35 MB