Statistics Toolbox™ 6.2

Latest Features

Version 6.2

Released: 01 Mar 2008
 

Version 6.2, part of Release 2008a, includes the following enhancements:

• New cross-validation functions for regression, classification, and clustering
• New sequential feature selection for regression and classification
• New quasi-random numbers for space-filling designs and Monte Carlo integration
• New partial least-squares regression for multicollinear data
• New nonnegative matrix factorization for dimension reduction

See the Release Notes for details.

Update Now

• Download the latest version if your software maintenance service is current.
• Purchase or request a renewal quote if your maintenance service has lapsed.
• Try the latest version by downloading or requesting trial software.

---

Version 6.1

Released: 01 Sep 2007
 

Version 6.1, part of Release 2007b, includes the following enhancements:

• Gaussian mixture distribution fitting and random number generation
• Copula parameter estimation from data
• Diagonal covariance matrix support for multivariate normal regression
• Enhanced boxplot style suitable for displaying a large number of groups
• Conversion from covariance matrix to correlation matrix

See the Release Notes for details.

Version 6.0

Released: 01 Mar 2007
 

Statistics Toolbox 6.0

• New dataset arrays for convenient organization and analysis of heterogeneous data and metadata
  • Encapsulate variables of different data types in a single array
  • Describe variables and observations using metadata, and access them by name
  • Create, manage, and operate on dataset arrays
• New categorical arrays for organizing and processing categorical data
  • Store both nominal (unordered) and ordinal data
  • Create, manage, and operate on categorical arrays
  • Use categorical arrays as input arguments in all Statistics Toolbox functions that use grouping variables
• New distribution fitting functionality for modeling outlier-prone data using empirical and Pareto models
• scatterhist function for producing scatter plots of 2-D data with marginal distributions along the axes
• linhyptest function for linear hypothesis testing of regression coefficients
• Classification and regression trees for producing objects to analyze model results
• Enhanced classify function for computing coefficients of discriminant functions defining boundaries between classification regions

See the Release Notes for details.

Version 5.3

Released: 01 Sep 2006
 
Version 5.3, part of Release 2006b, includes the following enhancements:

• New visualization capabilities -- main-effects plots, two-factor interaction plots, and multivariate charts for Design of Experiment results
• Expanded statistical process control functionality -- gage repeatability and reproducibility studies, eight new control charts, control rules for control chart data, and enhanced capability analysis functions
• New regression functionality -- multivariate regressions with missing data and multinomial regressions of discrete multicategory data, including multinomial logistic regression
• Expanded probability distributions -- multinomial PDF and random number generator
• Enhanced hypothesis testing -- new function to calculate sample size necessary for a hypothesis test to have a specified power, and improvements to Lilliefors and Jarque-Bera distribution tests to accommodate a wider range of sample sizes, significance levels, and probability distributions

See the Release Notes for details.

Version 5.2

Released: 01 Mar 2006
 

Version 5.2, part of Release 2006a, includes the following enhancements:

• Monte Carlo methods, including improved bootstrap capability and Markov Chain Monte Carlo tools
• ANOVA support for nested and continuous factors
• Expanded probability distributions: multivariate CDFs and more random number generators
• DOE improvements in D-optimal design generation and in generator selection for fractional factorial designs
• Copula enhancements and robust nonlinear regression

See the Release Notes for details.

Version 5.1

Released: 01 Sep 2005
 

Version 5.1, part of Release 14 with Service Pack 3, includes the following enhancements:

• New univariate probability distributions: generalized extreme value and generalized Pareto
• New hypothesis tests: chi-square goodness of fit, variance testing (including the Ansari Bradley and Levene tests), and a runs test of randomness
• New partial correlation functionality (linear and rank)
• New and enhanced functionality for prediction, inverse prediction, and prediction intervals
• New Cox proportional hazards regression for survival data
• Enhanced plotting usability for ksdensity and ecdf

See the Release Notes for details.

Version 5.0.2

Released: 07 Mar 2005
 

Version 5.0.2, part of Release 14 with Service Pack 2, includes bug fixes. See the Release Notes for details.

Version 5.0.1

Released: 22 Sep 2004
 

Version 5.0.1 includes bug fixes.

Version 5

Released: 02 Jun 2004
 

New features include:

• Distribution-fitting graphical user interface and command-line functions for:
  • Fitting and plotting distributions to data
  • Evaluating distributions at specific point
  • Handling censored (lifetime) data
• Support for N-Dimensional arrays in random number generation and descriptive statistics
• Several new multivariate plotting functions
• New functions and enhancements to maximum likelihood estimates (MLE) functionality
• Improved multivariate analysis capabilities, including:
  • Non-metric multidimensional scaling
  • More rotation options for factor analysis and pca
  • Faster and more flexible hierarchical clustering
• Faster exact p-values for nonparametric tests and control over approximate s exact calculation
• Rank correlation (Spearman and Kendall)

Version 4.1

Released: 22 Sep 2003
 

Version 4.1 includes a variety of new functions for:

• Analyzing hidden Markov models
• Analyzing Weibull distributions
• Analyzing extreme value distributions
• Stepwise regression and plotting
• Creating and editing the options structure of distribution fitting functions

Enhancements include:

• Improved graphical user interfaces for stepwise regression, plotting and generating distributions, robust fitting, and regression diagnostics for linear models
• Improved N-way analysis of variance for which you can now specify that grouping variables in an ANOVA model are random effects, using the syntax
• Distribution functions that now accept censored data