Dynamic Analysis with 1dCorrelate

AFNI Message Board
1

Hi all,

I am trying to do some dynamic analysis on my resting data, I extract the mean value of my ROIs in different time window and then calculate their Correlation with 1dCorrelate like this:
3dmaskave -mask $subj.r$run.PCC+tlrc. -q errts.$subj.r$run.e2t+tlrc.'[0-49]' > r$run.dy1.PCC.1D
3dmaskave -mask $subj.r$run.LmPFC+tlrc. -q errts.$subj.r$run.e2t+tlrc.'[0-49]' > r$run.dy1.LmPFC.1D
1dCorrelate -p r$run.dy1.PCC.1D r$run.dy1.LmPFC.1D
The ROIs I chosen are widely be defined as part of the default-mode network so they should be highly positive correlate with each other, but the output is zero, I wonder if I use this script wrong?

Best regards,
Sue

2

Without seeing the data, it is hard to say.
Personally, I would first look at the two .1D files – are the numbers “reasonable”?
Then, I would plot them, as in

1dplot -noline -x r$run.dy1.PCC.1D r$run.dy1.LmPFC.1D
and
1dplot -one -nopush r$run.dy1.PCC.1D r$run.dy1.LmPFC.1D

and see how these plots look. Depending on what I saw, then I’d proceed further.

Also, when you say something like “the output is zero”, it is MUCH MUCH better if you give the actual output. Was it EXACTLY zero, or just close to zero? I can’t tell from your posting what really happened. It’s good that you gave the exact commands you used. You should give the exact output as well.

3

Hi,

Sorry I didn't describe my problem clearly and didn't give the actual results. Here is my output:
y:/media/y/1E9042E09042BDD1/Resting/y02/y02.results> 1dCorrelate r$run.dy1.PCC.1D r$run.dy1.LmFC.1D
# Pearson correlation [n=50 #col=2]
# Name                Name                 Value   BiasCorr   2.50%   97.50%  N: 2.50% N:97.50%
# ------------------  ------------------  -------- -------- -------- -------- -------- --------
r01.dy1.PCC.1D[0]   r01.dy1.LmFC.1D[0]  +0.00000 +0.00000 -1.00000 +1.00000 -0.27835 +0.27835

I checked my 1Dfile with 1dplot:
1dplot -one r01.dy1.PCC.1D r01.dy1.LmFC.1D
Their curves TOTALLY overlapped, sorry I'm not sure how to upload the picture. 
http://http://photo.blog.sina.com.cn/photo/5769654217/006isSCZzy7e34LqHMbb7#pic

Best regards,
Sue

4

I have solved the problem, thank you anyway!!!

Best regards,
Sue

5

We like self-healing users! Very good.

6

Can I ask what you did? I just ran into the same issue where correlating the same two time series gives me a value of 0 on 1dCorrelate.

7

In my case, I finally found that my 2 ROIs were named the same in the beginning… :frowning:
It’s such a stupid problem but took me a few days to discover， an unforgettable experience.
Sorry for replying too late, hope you have solved your problem.

Sue