Department of Economics Tripolis Campus 22 100, Tel: +30-2170-230128 (Department),
2710-230132 Öáî: +30-2710-230139 |
|
Dimitrios D. Thomakos Associate Professor In Applied
Econometrics |
Scientific
Software and Programming
I am an avid fun of all kinds of statistical and
econometric software programs and I occasionally collect and try free and
shareware programs. Below you will find a list of my favorite programs for
doing most of the usual and not-so-usual statistical analysis…
Commercially available
O-Matrix (www.omatrix.com): my favorite programming
environment. O-Matrix is the fastest
matrix-based computation package in recent tests, is very user friendly and
allows for fast prototyping of complex algorithms and easy bookkeeping. I do
most of my advanced work using O-Matrix and I have written the toolbox Statistical
Time Series Analysis
currently in version 2.0.
EViews (www.eviews.com)
is a well-known favorite for doing standard and advanced applied econometric
work and producing nice graphs easily.
Kyplot (www.kyence.com) is (not well known but)
probably the premier spreadsheet-based statistical software package that I am
familiar with. It has an Excel-like GUI, it is extremely powerful (both in
quantity of commands and breadth of statistical applications), has excellent
graphing capabilities and one can do a presentation right from it. You can
download a free version of it at http://freestatistics.altervista.org.
S-Plus
(www.splus.com) is also well known and does the (few
other) things that the above programs cannot do (easily)…In contrast to the
above programs though S-Plus is relatively expensive to acquire as an
individual. Alternatively you can use R (www.r-project.org)
the free S-Plus clone!
Freeware (for
a comprehensive list of free statistical software visit http://freestatistics.altervista.org;
also try the online, web-based statistical calculation collection at John Pezzullo’s website http://members.aol.com/johnp71/javastat.html)
Dataplot (http://www.itl.nist.gov/div898/software/dataplot/homepage.htm)
is a free program for statistical analysis created, provided and updated by the
National Institute of Standards and Technology. It offers both a command-line
and a GUI version and it can handle a wide variety of complex statistical
analyses and graphing. The GUI version is user-friendly, after one masters a
few basic things. Its worthwhile the trouble to
download (you will need four different executables for the program to run
correctly).
MacAnova (http://www.stat.umn.edu/macanova)
is a free program for statistical analysis and matrix algebra. Older versions
were command-driven but it now features a GUI that improved a lot its easy of
use. You can easily import and export data and perform a lot of complex
statistical analyses using MacAnova and is worthwhile
to download and install. It’s also extensible with macros but that would
require that you go deeper in its programming language.
OpenStat 4 (http://www.statpages.org/miller/openstat)
is a free program for standard statistical analyses. It is very user-friendly,
allows for easy input of your data and produces text results that you can
easily incorporate in any word processor.
VisiCube (www.datamology.com)
is probably one of the best-designed programs for the visual and exploratory
analysis of quantitative data. Its statistical capabilities are very limited
but the underlying philosophy of data exploration and presentation is really
unique. You need to be patient to appreciate this program, especially in
preparing your data for reading and analysis.
For classroom use in introductory and
intermediate statistics courses
Stattucino (www.stattucino.com) is a great
easy-to-use program for use in the classroom. It does all kinds of analysis
expected for a sequence of two statistics courses at the above levels. You
don’t even have to download it! Just go online and use the Jave
applet that does exactly the same things that the desktop version will do. You
need to have your data online though for this choice (or, for small datasets
typical for classroom use, type them directly).
Qmulate (www.qmulate.co.zw) is currently in a
development stage and will probably go commercial soon. The current version is
freeware and worth the effort downloading and installing it. Very easy to use.
From the download you will only need the data manager.
Winstats (http://math.exeter.edu/rparris/winstats.html)
is a nice little program that does all the needed things for introductory
statistics students plus some nice simulations.