Free statistics software for Macintosh computers (Macs)
Updated 1-26-2010. Also see:
- Graphing and visualization software (including packages that do statistical routines, e.g. Aabel)
- Data mining, econometric software, general math, and specialty statistics.
- Dead software.
- Non-free statistics software for Macintosh computers
G*Power (free)
Configurations Available: 680x0; 680x0+FPU; PowerPC native; OS X (Universal Binary); Windows and DOS
Current Version: 3.0
Price: Free
Listing updated: 2/2007
The G*Power software is a freeware package developed by Axel Buchner of the Institut für Experimentelle Psychologie, Heinrich-Heine-Universität. It has its own web page and online manual. Their Web site says:
G*Power performs high-precision statistical power analyses for the most common statistical tests in behavioral research, that is, t-tests... F-tests (ANOVAS, multiple correlation and regression, and any other F-test), and Chi2-tests (goodness of fit and contingency tables). G*Power computes power values for given sample sizes, effect sizes, and alpha levels (post hoc power analyses), sample sizes for given effect sizes, alpha levels, and power values (a priori power analyses), and alpha and beta values for given sample sizes, effect sizes, and beta/alpha ratios (compromise power analyses).
Steve Hyde (John Radcliffe Hospital, University of Oxford) noted that the program is very stable and powerful, and that the new interface is a great improvement over the past versions.
gretl
gretl can do some general statistical routines and many specialized ones; it is in our “special purpose and general math programs” page.
GMT - The Generic Mapping Tools 
Configurations: UNIX, OS X; since you compile it, Universal
Current version: 4.51
Program updated: 9-20-09
Listing updated: 12-17-09
A collection of command-line tools that run on all Unix-like systems, including Mac OS X. See gmt.soest.hawaii.edu for details. Many of the main developers (including me) use Mac OS X. (Description by Paul Wessel)
Graphviz
Configurations: PPC
Current Version: 1.13
Listing updated: 12-17-09
Graphviz is the AT&T open source drawing package. The Mac OS X version and the overall project have their own web sites. The OS X version now uses the Aqua user interface. Prepare for a steep learning curve but it may be worth it if you have graphs you do frequently; not what I'd suggest for the occasional one-off though.
gnuplot
Configurations: PPC
Current Version: 4.2.6
Listing updated: 12-17-09
Program updated: 3-9-09 (version 4.4.0 is due soon)
Price: free
gnuplot is open source scientific plotting software. It is available online from many sources
MacAnova
Configurations Available: 68000; 680x0 + FPU; PowerPC native; G3 and G4 specific; OS X native; Universal
Current Version: 5.05
Listing updated: 4/24/08
MacAnova is a free, noncommercial, interactive statistical analysis program developed by Gary Oehlert and Christopher Bingham of the University of Minnesota School of Statistics. Their web site notes:
MacAnova has many capabilities but its strengths are analysis of variance and related models, matrix algebra, time series analysis (time and frequency domain), and (to a lesser extent) uni- and multi-variate exploratory statistics. MacAnova has a functional/command oriented interface. The Macintosh and Windows versions also have several window/menu/mouse type features. Although the language and syntax are S-like, MacAnova is not S or R.
MacAnova is now Intel native, and there is source code available. We found the program started up very quickly on a Intel Mini and had a fairly good menu system, which output visible code that we could copy and manipulate, or save and run later. It is almost similar to SPSS 4 in that regard, though better integrated into the system and lacking a separate output window. This is certainly worth a download. MacANOVA includes linear model and GLM routines.
Matplotlib
Free - open source - for Mac OS X
Currently in beta
Report updated: 5/15/07
Matplotlib is a pure python plotting library with the goal of making publication quality plots using a syntax familiar to matlab users. The library uses Numeric for handling large data sets and supports a variety of output backend
MX (free)
Mac OS X, Linux, many other versions
Current version: 1.55 (parity)
Last revised: April 19, 2004 (listing updated 6/2008)
Mx is a very cross-platform package that even has an on-line and Unix server version; it is written by Michael Neal. While there is a graphic user interface for Windows, the writer recommends doing a server installation to allow for a graphic interface. (Thanks, Stephen J. Read).
MegaStat
Current Version: 11.1
Configurations Available: requires Excel; works on Intel, PPC, Windows
Listing updated: 5-18-09
Cost: included with textbooks; download
MegaStat is maintained by J. B. Orris, Butler University, and distributed by McGraw-Hill, which explains why you may not have heard of it. The software uses Excel only for “data entry, data transformation, printing, and file management,” but avoids using Excel’s disreputable math tools.
MegaStat can deal with stepwise regressions, large factorials, time series/forecasting, descriptives, frequencies, nonparametrics, QPC sharts, and numerous hypothesis tests. In short, MegaStat packs all the power most people will ever need into a relatively inexpensive, easy to use package. The down side is that it’s moderately slow, has no scripting language, and requires Excel; and you may need to buy a textbook to get it, though J.B. Orris is considering a shareware or commercial version. On the Mac, some of the buttons and dialogues are hidden by formatting problems.
We have tested MegaStat’s output and in our tests, it was identical to six decimal points with that of Stata.
The software has evolved from J.B. Orris’ Microstat; the current version is written in VisualBasic as an Excel plugin, though a standalone version is planned for the distant future.
OpenEpi
Current Version: 2.2.1
Configurations Available: any platform with a Web browser capable of dealing with Javascript
Listing updated: 10-14-08
Kevin Sullivan’s open source OpenEpi software is available in four languages; unlike most software, it can be run from a web server or on a regular computer. The programs are written in Javascript and html and should be compatible with Macs and Linux and Windows machines. Test results are provided for each module to allow people to check reliability of their own setup. The software is set up for epidemiology and has numerous key statistics for that field, along with the usual means, medians, t-tests, ANOVAs, powers, etc.
PlainStat
Current Version: beta
Configurations Available: OS X, PPC or Intel
Listing updated: 11-13-08
Free software available at plainstat.com — it can do correlations, frequencies, descriptives, reliability, various t-tests, ANOVA, linear and canonical regression, factor analysis, K-means clustering, chi-squares, and various nonparametric tests. Plainstat is created and designed by Iyus A. Muslimin, statistician and Mac developer, who lived in Bandung, Indonesia.
Data is hard to type into the spreadsheet-style database at this point, but can easily be copied from spreadsheets such as Excel. Analysis is quick for small data sets and appears in an easy to read output bin in the main (and only) window; each analysis is added to a "table of contents" on the left side, along with the data. Output can be copied, but only without the column headings; it can also be exported to a tab-delimited plain text file for import into spreadsheets (or word processors, but since decimal points are fixed, that might not look very good). Data and value labels can be added via an inspector, a rare find in a free, simple program.
PlainStat might really be all you need — keeping in mind we have not yet verified its results by checking against a more established package. We're waiting with baited breath for the first release, which we hope will also have a facility for setting value and variable labels en masse.
PSPP
PSPP is a free SPSS clone. We are still investigating. “PSPP is a free program for statistical analysis of sampled data. It is similar to SPSS with a few exceptions. PSPP is particularly aimed at statisticians, social scientists and students requiring fast convenient analysis of sampled data.”
R (two entries)
There are two packages of R; the better known one is (CRAN).
R (CRAN) / “R for Mac OS X” / R.App and R GUI
Configurations Available: PowerPC (thru 1.7); Universal Binary (2.3+); Linux
Current Version: 2.9; under active development
Price: Free
Listing updated 5/4/09
This is an exceedingly flexible program, though hard to get started with; it is powerful and optimized for both PowerPC and Intel. A large number of libraries and built in routines are available. R will run many (if not most) S or S-Plus programs. The R programs and supplemental algorithms are distributed by the Comprehensive R Archive Network (CRAN). R will run easily on MacTel machines and loads very quickly.
A simple graphic user interface is now included for Mac users. The R Commander interface can be installed using the built-in package installer (check Install Dependencies); you may also want to use the package manager to install the various file import features (which aren't installed by default). There is also a Java interface, JGR, designed for the Mac. R Commander is an X11 program, which means it uses an alien interface and has odd open/save dialogues, but if you get past that it offers menu driven commands not dissimilar from, say, SPSS, just a bit more awkward ot use. Well, a lot more awkward, because you can't see the output window and have no spreadsheet data view!
R is similar to Bell Labs’ S system. It has a massive range of tests and now has Matrix as a recommended package, a useKerning argument for PDF and PostScript output, a recursive argument for file.copy(), an unzip function to expand or list zip archives, and other changes. For much more information about R, including advantages, drawbacks, resources, and tips, see our R statistics software for the Mac page.
Note: ADE-4 is a free, noncommercial, interactive statistical analysis program developed by Laboratoire de Biométrie, Génétique et Biologie des Populations in France. While the ADE-4 standalone program has been abandoned, a plug-in for R is available, free, and actively updated (the last update we noted was in April 2009).
R (Univ. Montreal)
Configurations Available: 680x0 (version 3); PowerPC/OS X (version 4); could probably be compiled for Intel
Current Version: 3 / Developer Release 4d10
Price: Free
Listing updated: 3/2008 (software last revised 4/2007)
This R is a full-featured public domain software package developed by the University of Montreal. It is only available for Mac and VAX/VMS (click here to read more or download it). Version 4.0 is (as of February 2006) still under development by Philippe Casgrain and “developer” versions are being freely distributed to a wide number of sites.
Regress+
Configurations Available: 68020 with FPU; PowerPC native; OS X (Carbon)
Current Version: 2.5
Price: Free
Listing updated: 8/07
Michael McLaughlin’s Regress+ is a free package that includes regression, stochastic modeling, bootstrapping and robust goodness of fit measures. The software and a tutorial are available at the Regress+ web site.
Regress 2.31 is available for older platforms, while version 2.5 is available for OS X and 9.2. Regress+ 3.0 is under development, according to Michael, who wrote: "it models distributions as well as equations and utilizes bootstrap techniques for goodness-of-fit tests and confidence intervals. It handles univariate models only." This program appears to cover every aspect of regression you can think of. It's graphically oriented but has strong statistics.
Good news from Michael McLaughlin:
Regress+ 3.0 is definitely under development although it is going very slowly. It will definitely be a Universal Binary and will be much better than the current v2.5.3, especially wrt speed. Tests show that v3.0 will be about 50x faster PER CPU (i.e.,. 100x for a dual-CPU Mac) compared to v2.5.3 on the same platform. Output graphics are a lot better, too — publication quality.
Unfortunately, I also have a day job which takes up almost all of my time. Added to that is what I am learning about new techniques and new algorithms, e.g., as of two weeks ago, I decided to change the PRNG (again!) to the new SFMT algorithm, replacing Mersenne Twister, since SFMT can use Altivec (or SSE2) instead of the main CPU.
StatCrunch
StatCrunch is fairly unique - a full-fledged statistics program that is freely available for web-based use, currently without advertisements, with a $5 per user fee for use on your own server, or $5/six months. It has the usual range of basic statistics, from t-tests to regression to ANOVA and nonparametric tests, with a wide range of graphs also available, and works from Excel or text files. StatCrunch, though free, will also store your data within reason; it seems to work fine with Safari. For those with low budgets or infrequent needs, StatCrunch's fairly easy to use interface and price are extremely attractive (it also makes sharing data easy).
Statistics101
Configurations: Java; should work on Intel and PPC Macs
Pricing: Free
Current Version: 1.06
Listing updated: 8/2007
Statistics101 is a giftware program designed to help learn probability and statistics the easy way—by simulation. “Gain deeper understanding of traditional statistics concepts and methods. Increase your awareness of the role of variability in probability and statistics. Learn and apply simple to very sophisticated statistical techniques without tables or complicated formulas.” Java program that runs cross-platform. Interprets and executes the simple “Resampling Stats” programming language. The original Resampling Stats language and computer program were developed by Dr. Julian Simon and Peter Bruce as a new way to teach Statistics to social science students.
SciPy
For Linux and Windows but may be compiled on Macs
SciPy is a library of scientific tools for Python which supplements the Numeric module. SciPy includes modules for graphics and plotting, optimization, integration, special functions, signal and image processing, genetic algorithms, ODE solvers, and others.
VTK (Visualization Toolkit)
Pre-compiled only for Windows; may be compiled from source code for OS X
Latest version: 4.2 (listing updated 8/2005)
The Visualization ToolKit (VTK) is a system for 3D computer graphics, image processing, and visualization with several interface layers. In VTK applications can be written directly in C++, Tcl, Java, or Python.
VTK supports a wide variety of visualization algorithms including scalar, vector, tensor, texture, and volumetric methods; and advanced modeling techniques like implicit modelling, polygon reduction, mesh smoothing, cutting, contouring, and Delaunay triangulation. Moreover, we have directly integrated dozens of imaging algorithms into the system so you can mix 2D imaging / 3D graphics algorithms and data. Our goal is to make the software easy enough for any computer literate person to use.
Vvidget
Configurations: Universal, 32-bit and 64-bit versions
Current Version: 10.5.5
Price: Free
Publisher/site: VVidget.org
Lance Bland, the developer, wrote: "In addition to all the standard features, Vvidget includes advanced features such as floating ticks, curves that can extend beyond the graph frame or can be truncated and literally hundreds of tunable parameters. 3D types rotate in real time and even the graph labels can be rotated in their own plane, independent of the main graph rotation. Data can be inserted through a list of numbers or by point and click methods." The same developer provides software such as QuadraticLab for other math functions. The software appears to be under very active development.