Skip to content

Latest commit

 

History

566 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

mechababs

Run BIDS Apps over many studies on an HPC cluster, with BABS running the jobs.

mechababs works on a BIDS study, adding derivatives to it, and records how each one was made inside the study itself: the exact mechababs and babs that ran, pinned in a lock, and every orchestration step as a datalad run record. A campaign is one such run's recipe, your app and cluster configs plus that locked environment, and mechababs iterate moves the study toward what you declared, one pass at a time. A superstudy groups many studies so one campaign sweeps them all.

Quickstart

datalad clone https://github.com/OpenNeuroStudies/study-ds000001 && cd study-ds000001
uvx --from git+https://github.com/con/mechababs@main mechababs campaign init demo \
    --apps ~/config/SimBIDS-0.0.3.yaml --cluster ~/config/your-site.yaml \
    --babs https://github.com/PennLINC/babs.git@main --limit 2
source .mechababs/campaigns/demo/env.sh
mechababs add-dataset --sourcedata sourcedata/ds000001
mechababs iterate      # repeat until every cell is merged

The quickstart walks through it, configs included. New cluster? Put the prerequisites in place (installation.md), then validate your cluster config with mechababs test-cluster (tutorial).

Docs

Upstream

About

Automate BABs at scale and collect orchestration provenance

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages