tomodrgn cleanup

Purpose

Clean an analyzed train_vae output directory of various types of outputs. Any model outputs that have been analyzed (i.e., those for which an analyze.EPOCH or convergence.EPOCH directory exist) will not be removed.

Sample usage

The examples below are adapted from tomodrgn/testing/commandtest*.py, and rely on other outputs from commandtest.py to execute successfully.

# Warp v1 style inputs -- dry run
tomodrgn \
    cleanup \
    output/vae_classE_sim_zdim8 \
    --weights \
    --zfiles \
    --volumes \
    --test

# Warp v1 style inputs -- actually remove files
tomodrgn \
    cleanup \
    output/vae_classE_sim_zdim8 \
    --weights \
    --zfiles \
    --volumes \
    --test

# WarpTools style inputs -- dry run
tomodrgn \
    cleanup output/vae_warptools_70S_zdim8 \
    --weights \
    --zfiles \
    --volumes \
    --test

Arguments

usage: cleanup [-h] [--weights] [--zfiles] [--volumes] [--test] workdir

Positional Arguments

workdir

Training directory containing training outputs to be cleaned

Named Arguments

--weights

Remove weights.N.pkl files directly within training directory, excluding those with a matching analyze.N or convergence.N subfolder

Default: False

--zfiles

Remove z.N.pkl files within directly training directory, excluding those with a matching analyze.N or convergence.N subfolder

Default: False

--volumes

Remove *.mrc volumes recursively within training directory. Note that volumes can be regenerated with config.pkl, weights.pkl, and appropriate z file.

Default: False

--test

Don’t actually delete any files, just list the files that would be deleted

Default: False

Common next steps

• Share the resulting cleaned directory with collaborators, or upload to Zenodo (or another data sharing service) as a component of data availability for publications