Back to PLS Help

brain score plot for behaviour analysis
archived_post
Posted on 02/02/07 11:59:40
Number of posts: 100
archived_post posts:

Author: Erin Skinner (---.uwaterloo.ca)
Date:   11-06-06 10:16

Hi,

I would also like to re-create the brain score vs. behaviour data plot, but this time for a block fMRI study. I tried to type the function bar(lvcorrs(:,1));
into matlab as you suggested in the last post, but it said that lvcorrs was undefined function or variable.
What I would like is the numerical values of the behaviour and brain score for each individual subject to re-create the plot.
If I need to do this in matlab rather than the gui, please write detailed instructions since I am not very familiar with the matlab or PLS program.

Thanks,

Erin

Replies:

Re: brain score plot for behaviour analysis
archived_post
Posted on 02/02/07 12:00:29
Number of posts: 100
archived_post replies:

Author: Randy (---.rotman-baycrest.on.ca)
Date:   11-06-06 13:12

Both variables are stored in the results file, but you will need to load the file into MATLAB and access the fields from the command line:

in MATLAB (the >> is the MATLAB prompt,)

>>load myresults_BfMRIresult

(replace 'myresults" with the actual filename for your results file)

>>save brainscores.txt b_scores -ascii

>>save behavdata.txt behavdata -ascii


Once you exit MATLAB you will have two text files, brainscores.txt and behavdata.txt, that you can import into your favorite plotting package.

have fun

Randy


Re: brain score plot for behaviour analysis
archived_post
Posted on 02/02/07 12:00:51
Number of posts: 100
archived_post replies:

Author: Jimmy (---.rotman-baycrest.on.ca)
Date:   11-06-06 13:41

Just a small correction:

Rather than b_scores, which is used for Behavior PLS part of Multiblock PLS, brainscores is used for Behavior PLS.

So, it should be

>>save brainscores.txt brainscores -ascii


Re: brain score plot for behaviour analysis
archived_post
Posted on 02/02/07 12:06:06
Number of posts: 100
archived_post replies:


Dear Randy and Jimmy,

Thanks so much for your help. Just to clarify, the data is organized so that each subject in each latent variable is a row (organized s1lv1, s2lv1,... sxlv1, s1lv2, s2lv2,....sxlv2...etc.) and the columns are the different conditions?

Thanks,

Erin


Re: brain score plot for behaviour analysis
archived_post
Posted on 02/02/07 12:06:50
Number of posts: 100
archived_post replies:

Author: Jimmy (---.rotman-baycrest.on.ca)
Date:   11-06-06 14:06

Each column stands for an LV, and the sequence of rows should always be subject-in-condition-in-group. i.e.:

g1c1s1LV1 g1c1s1LV2
g1c1s2LV1 g1c1s2LV2
g1c2s1LV1 g1c2s1LV2
g1c2s2LV1 g1c2s2LV2
g2c1s1LV1 g2c1s1LV2
g2c1s2LV1 g2c1s2LV2
g2c2s1LV1 g2c2s1LV2
g2c2s2LV1 g2c2s2LV2


Erin Skinner wrote:

> Dear Randy and Jimmy,
>
> Thanks so much for your help. Just to clarify, the data is
> organized so that each subject in each latent variable is a row
> (organized s1lv1, s2lv1,... sxlv1, s1lv2,
> s2lv2,....sxlv2...etc.) and the columns are the different
> conditions?
>
> Thanks,
>
> Erin


Re: brain score plot for behaviour analysis
archived_post
Posted on 02/02/07 12:07:10
Number of posts: 100
archived_post replies:

Author: Erin Skinner (---.uwaterloo.ca)
Date:   11-06-06 17:24

Great. Thanks!

Erin



Login to reply to this topic.

Contact Us:

3560 Bathurst Street
Toronto, Ontario
Canada M6A 2E1
Phone: (416) 785-2500

Baycrest is an academic health sciences centre fully affiliated with the University of Toronto