search

brainlife / app-ensembletracking

This App creates a large set of candidate streamlines using an ensemble of algorithms and parameter values. All outputs will be then combined into a single track.tck output.
0 stars 1 forks source link

The output track.tck is 22.3GB #4

Closed soichih closed 7 years ago

soichih commented 7 years ago

I've switched the tracking app for old Brain-Life workflow for LiFE from app-tracking to app-ensembletracking, and ran it with the first sample data (STN.sub-FP.run01_fliprot_aligned_trilin http://purl.stanford.edu/cs392kv3054).

The output from the app was 22.3GB behemoth and when I ran LiFE on it, it timed out after 3 hours (it looks like it would have taken more than 15 hours?)

fireshot capture 5 - white matter tracts_ - https___brain-life org_majortracts_

The 22.3GB output is probably unsustainable for a large number of subjects. Is there something wrong with the app, or input data?

francopestilli commented 7 years ago

This must be an error. If you are using the older life code from GitHub/Franco Pestilli/life it is possible that .tck us the .mat fe structure that can be very large.

The code we should use us GitHub/brain-life/encode

@ccaiafa

Franco

sent via phone, pardon brevity and mistakes.

On Jul 27, 2017 3:33 AM, "Soichi Hayashi" notifications@github.com wrote:

I've switched the tracking app for old Brain-Life workflow for LiFE from app-tracking to app-ensembletracking, and ran it with the first sample data (STN.sub-FP.run01_fliprot_aligned_trilin http://purl.stanford.edu/ cs392kv3054).

The output from the app was 22.3GB behemoth and when I ran LiFE on it, it times out after 3 hours (it looks like it would have taken more than 15 hours?)

[image: fireshot capture 5 - white matter tracts_ - https___brain-life orgmajortracts] https://user-images.githubusercontent.com/923896/28650382-9dc0a53e-7249-11e7-9049-910bc5fe25e8.png

The 22.3GB output is probably unsustainable for a large number of subjects. Is there something wrong with the app, or input data?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/brain-life/app-ensembletracking/issues/4, or mute the thread https://github.com/notifications/unsubscribe-auth/ACBYcz_qedTgPRFgfIP8NaHP0PlTFLLXks5sR-jvgaJpZM4Okr2Y .

soichih commented 7 years ago

@francopestilli I should be using the latest version of LiFE from brain-life/encode repo. I did pull recently.

Are you saying that it's an error with LiFE and 22G track.tck output from ensemble is normal?

soichih commented 7 years ago

Sorry, I just realized that app-ensembletracking uses much less fiber/fiber_max count than app-tracking. I think this explains why it was generating such a large output. I've reconfigured the defaults, and resubmitted some tests.

francopestilli commented 7 years ago

Good!

Franco

sent via phone, pardon brevity and mistakes.

On Jul 27, 2017 3:11 PM, "Soichi Hayashi" notifications@github.com wrote:

Sorry, I just realized that app-ensembletracking uses much less fiber/fiber_max count. I think this explains why it was generating such a large output. I've reconfigured the defaults, and resubmitted some tests.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/brain-life/app-ensembletracking/issues/4#issuecomment-318357455, or mute the thread https://github.com/notifications/unsubscribe-auth/ACBYc6o6tu-GKp7AIepefMqZyDtLU_oiks5sSIybgaJpZM4Okr2Y .