Ask Your Question

Statistical checks for regression, covariance, etc?

asked 2010-08-29 12:23:54 -0500

ccanonc gravatar image

updated 2015-01-14 03:55:55 -0500

FrédéricC gravatar image

Suppose I have some input vector(s) of scientific data. Assuming I at least know which ones are completely independent, and have some other(s) of which dependence is uncertain; what are some good sage functions to throw at the vector(s) to check for basic statistical parity/significance?

I know this question is very broad, so it's a basic howto rhetorical question. A continuous or discrete first order relationship is a fine intro; though more complex scenarios tend to be interesting.

edit retag flag offensive close merge delete

1 answer

Sort by » oldest newest most voted

answered 2010-08-29 21:54:21 -0500


Depending on what you mean by "scientific data" the answer would be different. If they are in a finite set say {True, False} then you can make a chi^2 test. If they are infinite but can be ordinate then you can perform a Mann-Whitney (or Wilcoxon or Wilcoxon-Mann-Whitney) test...

In the two preceding cases you should use R (which is included in Sage) which is a toolbox for statistics. More precisely you can look at

sage: r.chisq_test?
sage: r.wilcox_test?

(the r. means that you will use R)

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools

1 follower


Asked: 2010-08-29 12:23:54 -0500

Seen: 351 times

Last updated: Aug 29 '10