search

bbfrederick / rapidtide

rapidtide - a suite of programs for doing time lag correlation analysis on fMRI data
Apache License 2.0
75 stars 14 forks source link

Issue finishing happy using external cardiacfile in happy V2.0alpha12 #59

Closed NigelCol closed 3 years ago

NigelCol commented 3 years ago

Describe the bug I'm using the happy function using the fredericklab/rapidtide_dev:v2.0alpha12 version. I want to use happy providing an external cardiac text file through --cardiacfile. The function stops and gives me an error 'data dimensions not compatible with existing dimensions'. The --cardiacfreq and --cardiacstart argument values were taken from a physlog.json file. The cardiacfreq is 496 which is the frequency the PPU is sampled at for wireless philips scanners.

To Reproduce

docker run \
    --volume=INPUTDIRECTORY:/data_in,OUTPUTDIRECTORY:/data_out \
    fredericklab/rapidtide_dev:v2.0alpha12 \
        happy \
            /data_in/sub-08_task-rest_run-MB6_bold.nii.gz \
            /data_in/slice_timing.txt \
            /data_out/outputname \
           --cardiacfile=/data_in/cardiac_series.txt \
           --cardiacfreq=496
           --cardiacstart=-69.68548387096774

Error:

data dimensions not compatible with existing dimensions
25.0 25.0
0.0 0.0
['pleth'] ['cardiacfromfmri_25.0Hz', 'normcardiac_25.0Hz', 'envelope_25.0Hz', 'normcardiacfromfmri_dlfiltered25.0Hz', 'cardiacfromfmri_dlfiltered25.0Hz']
(5, 6373) (1, 8120)

Expected behavior Expected happy to finish without an error.

Screenshots If applicable, add screenshots to help explain your problem. image

Desktop (please complete the following information): Desktop Ubuntu 16.04, running the docker image from rapidtide_dev:v2.0alpha12

Additional context Attatch the cardiac text file used for the --cardiacfile argument and the phsylog.json file were I took the values for the --cardiacfreq and --cardiacstart arguments. The fMRI nifti is too large to upload here, but it should be the same one as in the shared dropbox. happy_issue.zip

bbfrederick commented 3 years ago

This should now be fixed in 2.0alpha13. Please let me know if it is not.

NigelCol commented 3 years ago

Hi, still having the same issue with the same data, using the 2.0alpha13 dev happy version.

bbfrederick commented 3 years ago

Would it be possible to put the data somewhere where I could try this out? I'm having a hard time figuring out exactly what's going on here.

NigelCol commented 3 years ago

I uploaded the data to a repository that you can find here : https://github.com/NigelCol/rapidtide_test. The data is under the data folder, I used the --cardiacfile command on the .tsv file using 'cardiac' as column selection.

bbfrederick commented 3 years ago

Ok, I got the data. I'll see what I can learn.

By the way - I just noticed that you don't have a continuation character on "--cardiacfreq=496" line, so happy will never see the cardiac start argument.

EDIT: That's not the problem - I'm getting the same result you are. Just an observation.

bbfrederick commented 3 years ago

Ok, I'm pretty sure I've fixed it now in 2.0alpha14. It was something totally different than what I thought - you were the first person to use a cardiac regressor file that didn't start at time 0 - my code read in the start time, then ignored it :-).

The dataset you sent works now.

NigelCol commented 3 years ago

I guess someone has to be the first :-). Closing because the issue is fixed!

Thanks.