python -m wanshi.visualizations.multi_cohort_scatter.multi_cohort_scatter \
--data-path-internal="/home/marcogustav/Documents/projects/xyz/results/crossval" \
--data-paths-external="/home/marcogustav/Documents/projects/xyz/results/deploy/ext_cohort_1" \
--data-paths-external="/home/marcogustav/Documents/projects/xyz/results/deploy/ext_cohort_2" \
--title="Prediction of Biomarkers from H&E pathology slides from XYZ" \
--internal-cohort-name="int_cohort" \
--format="norm" \
--outpath="/home/marcogustav/Documents/projects/gecco/results/crossval/figures/" data-path-internal includes subdirectories with the name of --format - e.g. raw and norm. These subdirectories
contain directories for each target. Each of the target directories contains a directory for each fold with the
prediction.csv.
The same applies to each data-paths-external directory.