lisa-sousa
commented
2 months ago - remove print statements
- adjust fetaure importance plot -> merge both fruntions, global and local feature importance plotting, into ine plot
- adjust importance value computation from 1-p-value to normalized(-log10(p-value)). clip the p-values at 1e-50 to avoid log10(0) and then normalize by log10(1e-50)
- rerun notebooks with updated code