Utility package to run Cascaded Hierarchical Model (CHM) jobs on clusters.
For more information visit our wiki page
- createchmjob.py -- Creates a set of CHM jobs to process a set of images on various compute clusters (Gordon, Comet, & Rocce)
- checkchmjob.py -- Checks status of jobs generated by createchmjob.py script
- createchmtrainjob.py -- Creates a CHM train job, runnable on various compute clusters (Comet & Rocce), which teaches CHM how to segment a desired feature. The output from this job is a trained model that is needed as an input for createchmjob.py
- createchmimage.py -- Simple image conversion utility
- createprobmapoverlay.py -- Overlays semi-transparent colored probability maps over base image
- mergetiles.py -- Merges a set of image tiles in a directory into single image
- chmrunner.py -- Internal tool that runs a CHM task created by createchmjob.py
- mergetilerunner.py -- Internal tool that runs Merge task created by createchmjob.py
- argparse
- configparser
- Pillow
- CHM singularity image (not required to build this software, but is needed to run the jobs)
- Should work on Python 2.7 & 3+ on Linux distributions
pip install chmutil
# creates a job assuming images are in ./images and ./trainedmodel has # chm model createchmjob.py ./images ./trainedmodel myrun --cluster rocce
See LICENSE.txt
Please report them here
This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.