search

PanyiDong / InsurAutoML

AutoML in Insurance project.
MIT License
5 stars 0 forks source link

Add_xgboost #9

Closed PanyiDong closed 2 years ago

codecov-commenter commented 2 years ago

Codecov Report

Merging #9 (bd4e8e5) into master (33afcd6) will increase coverage by 11.22%. The diff coverage is 51.47%.

@@             Coverage Diff             @@
##           master       #9       +/-   ##
===========================================
+ Coverage   29.26%   40.48%   +11.22%     
===========================================
  Files          41       47        +6     
  Lines        3332     4021      +689     
===========================================
+ Hits          975     1628      +653     
- Misses       2357     2393       +36
Flag Coverage Δ
unittests 40.48% <51.47%> (+11.22%) :arrow_up:

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files Coverage Δ
My_AutoML/_balancing/_under_sampling.py 52.39% <0.00%> (+1.65%) :arrow_up:
...hyperparameters/_ray/_classifier_hyperparameter.py 100.00% <ø> (ø)
..._hyperparameters/_ray/_regressor_hyperparameter.py 100.00% <ø> (ø)
My_AutoML/_imputation/_multiple.py 32.65% <ø> (+14.28%) :arrow_up:
My_AutoML/_model/_lightgbm.py 18.51% <18.51%> (ø)
My_AutoML/_utils/_optimize.py 51.98% <57.14%> (+48.59%) :arrow_up:
My_AutoML/_hpo/_ML.py 66.41% <60.00%> (+52.34%) :arrow_up:
My_AutoML/_model/_xgboost.py 70.17% <70.17%> (ø)
My_AutoML/_hpo/_base.py 47.56% <80.95%> (+40.86%) :arrow_up:
My_AutoML/_model/__init__.py 77.77% <100.00%> (ø)
... and 18 more

:mega: Codecov can now indicate which changes are the most critical in Pull Requests. Learn more

PanyiDong commented 2 years ago

build_nn failed as neural network fail to converge. I currently have no idea on how to improve the stability, but to run the test one more time to avoid reporting infinity (the direct reason build_nn failed). The pipeline still works (as I tested), but there are possibilities the fitting will fail.