issues
search
HelmholtzAI-Consultants-Munich
/
fg-clustering
Explainability for Random Forest Models.
https://forest-guided-clustering.readthedocs.io/
MIT License
26
stars
7
forks
source link
Speed up
#15
Closed
hpelin
closed
1 year ago
hpelin
commented
1 year ago
pytest on speed-up passed. Changes:
bootstrap matrix function (njit)
proximity matrix function (njit)
kmedoids package instead of scikit learn extra
docstring for run() function adjusted for the kmedoids package
adjustments in the test files: no more init_clustering k-medoids++ as it is not available in kmedoids & test optimizer for regression was failing with k=3 so I changed it to k=2
setup changed to new versions
pytest on speed-up passed. Changes: