search

neuro-ml / deep_pipe

A collection of tools for deep learning experiments
MIT License
37 stars 9 forks source link

Stratify val_ids in stratified_train_val_test_split #72

Open kirpoly opened 1 year ago

kirpoly commented 1 year ago

Was it intentional to just randomly take ids for validation? I guess in case of stratified split val_ids should be resemblant to the whole dataset too.

codecov[bot] commented 1 year ago

Codecov Report

Merging #72 (c6b9336) into develop (419e15b) will decrease coverage by 0.02%. The diff coverage is 0.00%.

@@             Coverage Diff             @@
##           develop      #72      +/-   ##
===========================================
- Coverage    36.20%   36.19%   -0.02%     
===========================================
  Files           70       70              
  Lines         3190     3191       +1     
  Branches       572      572              
===========================================
  Hits          1155     1155              
- Misses        1989     1990       +1     
  Partials        46       46
Impacted Files Coverage Δ
dpipe/split/cv.py 0.00% <0.00%> (ø)

:mega: We’re building smart automated test selection to slash your CI/CD build times. Learn more