Pipe-tree: declative pipelines based on FileTrees
Creators
- 1. Wellcome Centre for Integrative Neuroimaging (WIN), Centre for Functional Magnetic Resonance, Imaging of the Brain (FMRIB), University of Oxford, UK
Description
Framework to build pipelines in a declerative manner based on FileTrees. Pipe-tree has the following features:
-
The directory tree containing the input, intermediate, and output files are defined separately from the pipeline in the convenient file-tree format
-
The pipeline writer only defines the individual recipes in regular python code (e.g., how to convert file A into file B or file B into C), which pipe-tree will stitch together into an acyclical graph
-
Users can run part of the pipeline
-
By default, the pipeline will not rerun any jobs for which the output files already exist
-
Jobs can be run locally in sequence as normal python functions or in parallel by submitting to a cluster queue (using fsl_sub) or by using dask.
Available resources:
- Gitlab repository: https://git.fmrib.ox.ac.uk/ndcn0236/pipe-tree
- Documentation: https://open.win.ox.ac.uk/pages/ndcn0236/pipe-tree/
Files
pipe-tree-v0.2.1.zip
Files
(28.8 kB)
Name | Size | Download all |
---|---|---|
md5:175591d618d17200803bae64199122fa
|
28.8 kB | Preview Download |