[prfanalyze-vista] Unrecognized variable name 'stimulus_diameter'. #97
Comments
|
Please can you attach the config file?
What version of the prfanalyze are you using?
cheers,
Gari
…On Thu, Jan 30, 2025 at 10:29 AM Gilles de Hollander < ***@***.***> wrote:
Hey guys,
I'm trying some packages to fit synthesized data from prfsynth.
I created data using
docker run --rm -it \
--platform linux/amd64 \
-v "$script_dir/$json_file:/flywheel/v0/input/config.json" \
-v "$basedir:/flywheel/v0/output" \
garikoitz/prfsynth
With the json-file basically the default-config of prfsynth.
I then get the following error when I try to fit:
./fit_mrvista.sh verysmall
Running with identifier: verysmall
Using config file: /home/ubuntu/git/paper_braincoder/prfsynth/fit/mrvista/configs/prfanalyze-vista-verysmall.json
WARNING: Using temporary output directory: /flywheel/v0/output/BIDS/derivatives/prfanalyze-vista_temp001
[base/run.py] Using real data, not coming from prfsynthesize.
[base/run.py] This is the temp file with stim info:
/tmp/tmpzu447l.json
[base/run.py] This is the content:
{u'fieldofviewHorz': 20, 'isPRFSynthData': False}
------------------------------------------
Setting up environment variables
---
LD_LIBRARY_PATH is .:/opt/mcr/v95//runtime/glnxa64:/opt/mcr/v95//bin/glnxa64:/opt/mcr/v95//sys/os/glnxa64:/opt/mcr/v95//sys/opengl/lib/glnxa64
This is the config.json file being read: /running/out/options.json
These are the contents of the json file:
tmp =
struct with fields:
keepAllPoints: 0
coarseDecimate: 2
prfprepareAnalysis: '01'
numberStimulusGridPoints: 50
detrend: 1
wsearch: 'coarse to fine'
grid: 0
model: 'one gaussian'
hrf: {0x1 cell}
================================================================================
/flywheel/v0/input/BIDS/sub-verysmall/ses-1/func/sub-verysmall_ses-1_task-prf_acq-normal_run-01_bold.nii.gz
/tmp/tmpzu447l.json
/flywheel/v0/input/BIDS/stimuli/sub-verysmall_ses-1_task-prf_apertures.nii.gz
/running/out/options.json
--------------------------------------------------------------------------------
This is the options file comming in:
allOptions =
struct with fields:
vista: [1x1 struct]
aprf: [1x1 struct]
afni: [1x1 struct]
mlr: [1x1 struct]
Error using pmModelFit (line 128)
Unrecognized variable name 'stimulus_diameter'.
Error in prfanalyze_vista (line 96)
MATLAB:table:UnrecognizedVarName
real 0m15.149s
user 0m12.074s
sys 0m2.320s
<ERROR> An error occurred during execution of the Matlab executable. Exiting!
Note that prfsynth/aprf, prfsynth/popeye, prfsynth/afni can all fit this
data out-of-the-box.
Cheers,
Gilles
—
Reply to this email directly, view it on GitHub
<#97>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABCZAV7ZK7LB2M4UV7QVUMD2NHWIXAVCNFSM6AAAAABWEZDWYSVHI2DSMVQWIX3LMV43ASLTON2WKOZSHAZDAMZRHAZTIOA>
.
You are receiving this because you are subscribed to this thread.Message
ID: ***@***.***>
|
|
Here you go. |
|
can you try adding stimulus_diameter to the prfanalyze config? can you
download the default for that gear?
are you using prfanalyze-vista:2.3.0_3.1.2?
thanks!
…On Thu, Jan 30, 2025 at 10:53 AM Gilles de Hollander < ***@***.***> wrote:
Here you go.
prfsynth-verysmall.json
<https://github.com/user-attachments/files/18600546/prfsynth-verysmall.json>
prfanalyze-vista-verysmall.json
<https://github.com/user-attachments/files/18600559/prfanalyze-vista-verysmall.json>
—
Reply to this email directly, view it on GitHub
<#97 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABCZAV2T6D3BFSVHLJ46VI32NHZAPAVCNFSM6AAAAABWEZDWYSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRUGAYTGNJVGY>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
|
Hey @garikoitz , Yes. That was the issue. When I use this config: That works! Note that the stimulus_diameter is missing from your default config file One more note: When I try to run your package as a singularity container, I get the following error: Again, it works fine with afni and popeye (but not This makes it impossible to run this stuff on our cluster. Is this a known issue? Do you want me to open a new one for that? |
|
weird! will test, we use Singularity too
…On Thu, Jan 30, 2025 at 11:20 AM Gilles de Hollander < ***@***.***> wrote:
Hey @garikoitz <https://github.com/garikoitz> ,
Yes. That was the issue. When I use this config:
{
"subjectName" : "verysmall",
"sessionName" : "1",
"solver" : "vista",
"tasks" : "[all]",
"isPRFSynthData" : true,
"options": {
"prfprepareAnalysis" : "01",
"hrf": [],
"model" : "one gaussian",
"grid" : false,
"wsearch" : "coarse to fine",
"detrend" : 1,
"keepAllPoints" : false,
"numberStimulusGridPoints": 50,
"coarseDecimate": 2
},
"stimulus": {
"fieldofviewHorz":20,
"stimulus_diameter":20
}
}
That works!
Note that the stimulus_diameter is missing from your default config file
<https://github.com/vistalab/PRFmodel/blob/51ab893548dac1e914ba93da7580e3fae020010b/gear/prfanalyze/vista/default_config.json#L18>
One more note: When I try to run your package as a singularity container,
I get the following error:
CONFIG_FILE=$PWD/configs/prfanalyze-vista-${IDENTIFIER}.json
OUTPUT_DIR=/shares/zne.uzh/gdehol/ds-prfsynth
SIF_IMAGE=/shares/zne.uzh/containers/prfanalyze-vista.sif
echo "Running with identifier: $IDENTIFIER"
echo "Using config file: $CONFIG_FILE"
# Run Singularity with writable-tmpfs
singularity exec --cleanenv --writable-tmpfs \
--bind "$OUTPUT_DIR:/flywheel/v0/input" \
--bind "$OUTPUT_DIR:/flywheel/v0/output" \
--bind $CONFIG_FILE:/flywheel/v0/input/config.json \
"$SIF_IMAGE" \
/flywheel/v0/run.sh
Running with identifier: verysmall
Using config file: /home/gdehol/git/paper_braincoder/prfsynth/fit/mrvista/configs/prfanalyze-vista-verysmall.json
INFO: Mounting image with FUSE.
Preparing solver "prfanalyze-vista":
[base/run.py] Using synthetic data coming from prfsynthesize.
Subject: verysmall
Session: 1
Options: {'prfprepareAnalysis': '01', 'hrf': [], 'model': 'one gaussian', 'grid': False, 'wsearch': 'coarse to fine', 'detrend': 1, 'keepAllPoints': False, 'numberStimulusGridPoints': 50, 'coarseDecimate': 2}
WARNING: Using temporary output directory: /flywheel/v0/output/BIDS/derivatives/prfanalyze-vista_temp006
Output BIDS directory: /flywheel/v0/output/BIDS/derivatives/prfanalyze-vista_temp006/sub-verysmall/ses-1
Beginning os.wait() for /solve.sh, run=01 (child pid is 3991966)
Failed to exec /solve.sh script!
No estimates.json file found.
Again, it works fine with afni and popeye (but not prfanalyze-aprf!)
This makes it impossible to run this stuff on our cluster. Is this a known
issue? Do you want me to open a new one for that?
—
Reply to this email directly, view it on GitHub
<#97 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABCZAV6NJPSBOXNRSNCI7ST2NH4FVAVCNFSM6AAAAABWEZDWYSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRUGA3TMMBUGU>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hey guys,
I'm trying some packages to fit synthesized data from prfsynth.
I created data using
With the json-file basically the default-config of prfsynth.
I then get the following error when I try to fit:
Note that
prfsynth/aprf,prfsynth/popeye,prfsynth/afnican all fit this data out-of-the-box.Cheers,
Gilles
The text was updated successfully, but these errors were encountered: