Regress Plus - A very user-friendly state-of-the-art tool for mathematical modeling.

Version:Regress+ is a user-friendly, state-of-the-art tool for mathematical modeling.2.5.3

License:Free

Operating System:Mac OS X

Homepage: www.causascientia.org

Developed by:

It will optimize almost any univariate deterministic model (equation) as well as stochastic models (distributions) having at most five parameters.

Regress Plus features and functionality cannot be matched by any other available program.

Here are some key features of "Regress Plus":

General

· Simple (univariate) mathematical modeling

· Data:

· deterministic (regression)

· stochastic (random variates)

· Up to 2,147,483,647 points (minimum 7)

· Robust goodness-of-fit testing

· Bootstrap confidence intervals (90, 95, and 99 percent) for parameters

· stochastic-model goodness-of-fit metrics

· Bootstrap methodology (where appropriate):

· BCa technique (state-of-the-art)

· percentile technique

· tunable precision

· Optional "freezing" of initial estimate(s) for any parameter(s)

· Choice of optimization criterion

· No hidden assumptions anywhere:

· no approximations, apart from those common to sampling and bootstrapping generally

· no data transformations of any kind

· "Smart" dialogs to run unattended and/or in the background

· Textfile input (may contain unlimited comments)

· Both text and graphical output

· One keystroke makes a plot (may be saved as a PICT)

· Extensive documentation (in PDF):

· Tutorial (50 pp.)

· Users' Guide (36 pp.)

· Technical Details and References (8 pp.)

· Appendix A: A Compendium of Common Probability Distributions (120 pp., separate volume)

· Appendix B: Error Messages (4 pp.)

· Lots of sample datafiles

Deterministic Modeling

· Models: y = f(x), with 1 to 10 parameters

· 22 Built-in families of models

· Gaussian-Lorentzian model for spectral peaks (see Example #5)

· "User-defined" model

· Special Simulated Annealing mode to help find initial parameter estimates

· Optimization criteria:

· Least-squares

· Minimum average deviation

· [Optional] Weights (for the dependent variable, y)

· [Optional] Listing of fitted data and residuals

Stochastic Modeling

· 56 Built-in distributions:

· 30 Continuous

· 17 Continuous binary mixtures

· 5 Discrete

· 4 Discrete binary mixtures

· Optimization criteria:

· Maximum-likelihood (all)

· Minimum Kolmogorov-Smirnov statistic (continuous variates)

· Minimum Chi-square (discrete variates)

· [Optional] Discrete input may be grouped

· [Optional] Creation of samples of random variates (textfiles, see Example #10)