comp.lang.idl-pvwave archive: archive

Home » Public Forums » archive » vector t-test?

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

Re: vector t-test? [message #36912 is a reply to message #36821]

Mon, 03 November 2003 11:16

mmiller3
Messages: 81
Registered: January 2002

Member

>>>> > "wmc" == wmc <wmc@bas.ac.uk> writes:

> Can I do a t-test on vectors (I mean, where each data point
> is a 2-d vector)? tm_test says it works on reals; does it
> work on complex data?

No - a t-test compares the means of two populations. If you have
multidimensional data, you can compare means along multiple axes.
Those may not be orthogonal though (even if the axis are
spatially orthogonal), so the results may not be independent.

To get a better measure of the difference between the two
populations, take a look at various spatial statistics methods to
find the "right" statistic for your use. I don't know what is
available in IDL though...

Regards, Mike

--
Michael A. Miller mmiller3@iupui.edu
Imaging Sciences, Department of Radiology, IU School of Medicine

Report message to a moderator

[Message index]

		vector t-test? By: wmconnolley on Mon, 03 November 2003 07:19
		Re: vector t-test? By: mmiller3 on Mon, 03 November 2003 11:16

Previous Topic:	How to make IDL faster
Next Topic:	Re: Keywords/Parameters and Common Blocks

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Fri Oct 10 08:59:46 PDT 2025

Total time taken to generate the page: 1.27998 seconds