ENHANCEMENT: Provide a complete functional pipeline usable with test data

[gdevenyi]

In order to ensure an installation of RABIES is functional, we need to exercise all steps of the pipeline with a minimal dataset.

That means

• data
• commands

This can be used to ensure that any particular commit on the code works.

Related to #301 as I don't know if my changes break anything.

[gdevenyi]

Example data runs with this:

#!/bin/bash

set -euo pipefail

docker run -it --rm -v $PWD/work_dir:/work_dir --user "$(id -u):$(id -g)" \
rabies/testing -p MultiProc \
preprocess /work_dir/test_dataset /work_dir/rabies_out \
--anatomical_resampling 0.3x0.3x0.3

docker run -it --rm -v $PWD/work_dir:/work_dir --user "$(id -u):$(id -g)"  \
rabies/testing -p MultiProc \
confound_correction /work_dir/rabies_out /work_dir/rabies_out

docker run -it --rm -v $PWD/work_dir:/work_dir --user "$(id -u):$(id -g)"  \
rabies/testing -p MultiProc \
analysis /work_dir/rabies_out /work_dir/rabies_out --data_diagnosis

Add to testing action

[Gab-D-G]

Here's a set of commands (from https://github.com/grandjeanlab/INCF_preclinical/blob/main/rabies/process_test_dataset.ipynb) that should work on the new test dataset (https://zenodo.org/record/8349029):

#!/bin/bash

set -euo pipefail

docker run -it --rm \
-v $PWD/test_dataset:/test_dataset:ro \
-v $PWD/test_dataset_outputs:/test_dataset_outputs/ \
gabdesgreg/rabies:0.4.8 -p MultiProc \
preprocess /test_dataset/ /test_dataset_outputs --anatomical_resampling 0.15x0.15x0.15 \
--anat_inho_cor method=SyN,otsu_thresh=2,multiotsu=true

docker run -it --rm \
-v $PWD/test_dataset:/test_dataset:ro \
-v $PWD/test_dataset_outputs:/test_dataset_outputs/ \
gabdesgreg/rabies:0.4.8 -p MultiProc --local_threads 4 \
confound_correction /test_dataset_outputs/ /test_dataset_outputs \
--frame_censoring FD_censoring=true,FD_threshold=0.05,DVARS_censoring=false,minimum_timepoint=80 \
--conf_list mot_6 WM_signal CSF_signal --smoothing_filter 0.3

docker run -it --rm \
-v $PWD/test_dataset:/test_dataset:ro \
-v $PWD/test_dataset_outputs:/test_dataset_outputs/ \
gabdesgreg/rabies:0.4.8 -p MultiProc --local_threads 4 \
analysis /test_dataset_outputs/ /test_dataset_outputs \
--data_diagnosis

QC outputs should reproduce what's found in https://github.com/grandjeanlab/INCF_preclinical/tree/main/rabies/test_dataset_QC.

What would be the next steps to implement this systematically?

[gdevenyi]

This will be added to the GitHub action as a testing workflow. I need to figure out if it goes with the existing build or a seperate thing. @david-roper might do this

[gdevenyi]

Example data and testing commands need to exercise every part of the pipeline.

Current container does not appear to properly run melodic due to lack of testing during development.

[Gab-D-G]

This is addressed #326. Essentially all functions not conducted with error_check_rabies.py will be complemented by processing real data, with the test dataset. This is now as complete as it can be at the moment.