Niche Modeling

8
Jul
Standard Statistical Packages

If you have an interest in Statistics packages, for finance, investment, or science, you soon find out how expensive they are. I thought I would do a short survey of the standard statistical packages, and see how the free ones compare.

MiniTab — “The World Trusts Minitab for Quality”. A 30 day free demo is available. The market for Minitab is corporate quality improvement. It appears to have a nice graphics and user interface and a simple macro language. The cost is $1195. The product seems relatively simple compared with the other statistics packages.

SAS — “The power to know”. SAS is very strong on integrated business analytics and prediction, with a comprehensive array of 100s of different products for different industries and applications. For example, integrated business systems include a database and visualization. There was no published price information.

SPSS — “Enabling the predictive enterprise”. SPSS is pitching predictive enterprise concepts aimed at increasing return on investment in the corporate environment. They have many different products across technologies and industries (e.g. risk management, fraud, data mining, healthcare) with each package about $400-$500 (e.g. classification and regression trees is a package).

S-plus — “Delivering the Knowledge to Act”. This is a package with a language based on S oriented towards industries e.g. life sciences, telecommunications, manufacturing (and not so corporate finance oriented). No price was provided.

Matlab — “Accelerating the pace of engineering and science”. Matlab is strong on numeric simulation in engineering fields, with its Matsim package, can integrate many complex applications in statistics and signal processing. The web site claims a 15-day trial is available for most products but only if have an existing product license. The cost is dependent on academic or other pricing, but expect a few hundred for basic package, and additional hundreds for each extra module.

Maple 10 is a little different from the other statistics packages, as an algebra system, it can manipulate equations, substitute, simplify, manage, and help publish the most complex mathematics. It also works with Matlab.

Free Packages

Here are the main free packages. I have used R for many years with no complaint, except the memory management is a bit flaky, but perhaps better coding style on my part by freeing objects would help.

R — “R Project for Statistical Computing” is a free software environment for statistical computing and graphics with many user contributed packages for many applications. The language is virtually identical to Splus and should be able to be used instead of Splus in many cases.

The following are emulators for Matlab in R. I have not used these.

R.matlab — “Read and write of MAT files together with R-to-Matlab connectivity” — This package provides methods to read and write MAT files. It also makes it possible to communicate (evaluate code, send and retrieve objects etc.) with Matlab v6 or higher running locally or on a remote host.

matlab — Package emulating Matlab routines not implemented in R.

There are also a number of free Matlab look-alikes:

Scilab — “A Free Scientific Software Package”. Excellent documentation, excellent support via e-mail and its own newsgroup, source code, and Windows binary available. Mostly compatible with Matlab.

Octave — GNU Octave is a free high-level statistics program language. Reportedly offers better compatibility with Matlab than Scilab but the Windows version might more complex to install.

rlabplusis — “A continuation of work on an open-source scripting environment for scientific computations”. Language inspired by but not compatible with Matlab. Numerical libraries and toolkits, e.g., GNU Scientific Library (GSL).

Euler is a powerful numerical laboratory with a programming language that comes with Yacas, a computer algebra system. This would be equivalent to a free GNU general license Maple plus Matlab, but not a clone.

My problem is to run the RegEM algorithm for recovering climate fields from simulations, supplied with the controversial paper “Testing the Fidelity of Methods Used in Proxy-based Reconstructions of Past Climate“, Journal of Climate, 18 by Mann et.al.. Based on my survey, I think I will first try the free program Octave to see if I can get RegEM to run. The R emulators are worth a look, but do not look as though they will do the job. GNU software is generally very efficient and reliable, once the compilation hurdle is overcome.

Tagged as:
Published by david stockwell in: All
If you like this blog please take a second from your precious time and subscribe to my rss feed!

Martin Ringo

David,

A large hunk of the kind of statistics that is used in climate analysis is essentially the same algorithms as used in econometrics — although view climatologists seem to cite econometric references. Of your list of statistical packages SAS is commonly used in econometrics and has an extensive set of procedures fully debugged and fairly easy to use. I would add to the list the following:

1) BMDP: a long-running series of general purpose statistical packages.
2) EViews: the old Mirco-TSP turned into a Windows program.
3) EasyReg: a free, non-commerical econometric package similar to EViews 3.1
(you can get EasyReg at http://econ.la.psu.edu/~hbierens/EASYREG.HTM)
4) GAUSS: the classic matrix programming language used in econometrics
5) JMP: a graphics-rich package for exploratory analysis.
6) Minitab: a general purpose statistical analysis package.
7) Microfit: for time series analysis.
RATS and CATS: both time-series oriented.
9) SHAZAM: widely used econometric software.
10) Soritech: a general purpose econometrics package
11) SPSS: very popular for use in applied statistics for social sciences.
12) Stata: one of the most popular econometrics packages. version 6.0 available.
13) SYSTAT: a popular general purpose statistical analysis package.
14) TSP: the classic time series analysis package.

I have used BMDP (what we used to call “BioMed”), EViews, GAUSS, Soritech, SPSS, SYSTAT and a few others not listed. I like GAUSS for modeling because the random number generators have pass the “Die Hard” pseudo random number generation tests and the matrix syntax is easy to master. Not quite APL but good enough … especially since I could even read APL code any more. However, none of the strict econometric packages have anything like the library that R has. For instance, the simulation of fractionally integrated pseudo random deviates is a simple call in R and a subroutine I haven’t quite written right in GAUSS.

That said, I would still rather use EViews, Rats, or Shazam for anything resembling the general linear model because of the easier access to standard model tests and the much better documentation of what is going on. The R help files don’t even always list the input choices and would never dream of actuall telling the user of what the underlying algorithms are. Just a reference to a paper, which if you don’t have handy, tough luck. What R needs is someone to write an “Econometric Foundations” (Mittlehammer, Judge and Miller) with an R implementation as the authors did for GAUSS and leaving all routines a procedures in source code. One can then learn the econometrics, modeling, and coding together though it is easier if you already know one of the three.

Anyway I suggest that you go visit Herman Bierens EasyReg site and download a free econometrics package just to get the feel of that kind of program.
Martin Ringo

David,

A large hunk of the kind of statistics that is used in climate analysis is essentially the same algorithms as used in econometrics — although view climatologists seem to cite econometric references. Of your list of statistical packages SAS is commonly used in econometrics and has an extensive set of procedures fully debugged and fairly easy to use. I would add to the list the following:

1) BMDP: a long-running series of general purpose statistical packages.
2) EViews: the old Mirco-TSP turned into a Windows program.
3) EasyReg: a free, non-commerical econometric package similar to EViews 3.1
(you can get EasyReg at http://econ.la.psu.edu/~hbierens/EASYREG.HTM)
4) GAUSS: the classic matrix programming language used in econometrics
5) JMP: a graphics-rich package for exploratory analysis.
6) Minitab: a general purpose statistical analysis package.
7) Microfit: for time series analysis. RATS and CATS: both time-series oriented.
9) SHAZAM: widely used econometric software.
10) Soritech: a general purpose econometrics package
11) SPSS: very popular for use in applied statistics for social sciences.
12) Stata: one of the most popular econometrics packages. version 6.0 available.
13) SYSTAT: a popular general purpose statistical analysis package.
14) TSP: the classic time series analysis package.

I have used BMDP (what we used to call “BioMed”), EViews, GAUSS, Soritech, SPSS, SYSTAT and a few others not listed. I like GAUSS for modeling because the random number generators have pass the “Die Hard” pseudo random number generation tests and the matrix syntax is easy to master. Not quite APL but good enough … especially since I could even read APL code any more. However, none of the strict econometric packages have anything like the library that R has. For instance, the simulation of fractionally integrated pseudo random deviates is a simple call in R and a subroutine I haven’t quite written right in GAUSS.

That said, I would still rather use EViews, Rats, or Shazam for anything resembling the general linear model because of the easier access to standard model tests and the much better documentation of what is going on. The R help files don’t even always list the input choices and would never dream of actuall telling the user of what the underlying algorithms are. Just a reference to a paper, which if you don’t have handy, tough luck. What R needs is someone to write an “Econometric Foundations” (Mittlehammer, Judge and Miller) with an R implementation as the authors did for GAUSS and leaving all routines a procedures in source code. One can then learn the econometrics, modeling, and coding together though it is easier if you already know one of the three.

Anyway I suggest that you go visit Herman Bierens EasyReg site and download a free econometrics package just to get the feel of that kind of program.
http://landshape.org David Stockwell

Martin,
Thanks for the additions. I looked around and couldn’t find much in the way of reviews, so these are all helpful. Other information apart from links and markets would be whether there are equivalences like free versions of commercial packages.

I will check the EasyReg site out.
http://landshape.org David Stockwell

Martin,
Thanks for the additions. I looked around and couldn’t find much in the way of reviews, so these are all helpful. Other information apart from links and markets would be whether there are equivalences like free versions of commercial packages.

I will check the EasyReg site out.

Standard Statistical Packages

Free Packages

Categories

Blogroll

Advertising

Comments

Archives

tag cloud

Most commented

Copyright © 2008 - 2012 Niche Modeling