statistical analysis tools in python?

[Thomas Nelson]

Sorry if this is a FAQ, but I couldn't find a good summary through
google. What kinds of statistical analysis tools exist in python? I
really just need t-tests, chi-squared test, and other such tests of
statistical significance. A few things point to numpy and scipy, but I
was surprised to find the documentation for numpy is not freely
available, and I thought it would be wise to ask here before I download
it and start hunting through the source code for what I want. Is numpy
the best option for my simple needs? Also I was a little surprised to
find nothing in the builtin python modules that can find standard
deviation, quartiles, etc. Is this for any particular reason, or
perhaps no one has shown any interest? I'd be willing to work on a
project to make simple single-variable analysis part of the builtin
python distribution.

Thanks all,

THN

 

[Robert Kern]

Sorry if this is a FAQ, but I couldn't find a good summary through
google. What kinds of statistical analysis tools exist in python?

The third hit for "python statistics" is a good overview of what's available:

http://www.astro.cornell.edu/staff/loredo/statpy/

I
really just need t-tests, chi-squared test, and other such tests of
statistical significance. A few things point to numpy and scipy, but I
was surprised to find the documentation for numpy is not freely
available,

So? The web page does give an overview of what numpy has, and the sample
chapters go into more depth.

and I thought it would be wise to ask here before I download
it and start hunting through the source code for what I want. Is numpy
the best option for my simple needs?

numpy tries to be an array package, not a stats package and thus has nothing for
statistical tests. However, scipy has plenty of stats functionality going well
beyond what you've specify.

--
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
that is made terrible by our own mad attempt to interpret it as though it had
an underlying truth."
-- Umberto Eco

 

[gblais]

And there is a python interface to R, so that you can call R routines from
Python. R is a free stat language that has all the tests you've mentioned,

Gerry

 

[Thomas Nelson]

Actually, after a little looking, the simple stats.py module at
http://www.nmr.mgh.harvard.edu/Neural_Systems_Group/gary/python.html
is exactly what I needed. It may not be as fast or as comprehensive as
scipy or R, but installation simply involves downloading the module and
importing into the code, and it has the ttests I was looking for.
Thanks,
THN