FEA allow any resampler in the BalancedBaggingClassifier

[glemaitre]

closes #653

It allows implementing the following methods just by swapping the sampling:

• Over-Bagging
• Under-Bagging
• Under-Over-Bagging
• SMOTE-Bagging

TODO:

• Add more details in documentation
• Add reference to articles
• Add an example
• Add a small section in the user guide.
• Add a test to ensure the support for FunctionSampler by bypassing the sampling_strategy validation.

[glemaitre]

I think that we need to be careful with the bootstrap option to see if we don't draw twice some bootstrap samples.

[codecov]

Codecov Report

Merging #808 (d1d9546) into master (76abfa4) will decrease coverage by 2.78%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #808      +/-   ##
==========================================
- Coverage   98.59%   95.80%   -2.79%     
==========================================
  Files          93       91       -2     
  Lines        6045     5959      -86     
  Branches      503      504       +1     
==========================================
- Hits         5960     5709     -251     
- Misses         84      192     +108     
- Partials        1       58      +57     

Impacted Files	Coverage Δ
imblearn/ensemble/_bagging.py	96.15% <100.00%> (-1.68%)	⬇️
imblearn/ensemble/tests/test_bagging.py	100.00% <100.00%> (ø)
imblearn/keras/tests/test_generator.py	7.93% <0.00%> (-84.13%)	⬇️
imblearn/keras/_generator.py	55.71% <0.00%> (-41.43%)	⬇️
imblearn/tensorflow/tests/test_generator.py	65.59% <0.00%> (-34.41%)	⬇️
imblearn/utils/deprecation.py	80.00% <0.00%> (-20.00%)	⬇️
imblearn/tensorflow/_generator.py	83.87% <0.00%> (-16.13%)	⬇️
imblearn/datasets/_imbalance.py	76.47% <0.00%> (-11.77%)	⬇️
imblearn/over_sampling/_smote/filter.py	86.44% <0.00%> (-6.78%)	⬇️
imblearn/ensemble/_weight_boosting.py	92.13% <0.00%> (-5.62%)	⬇️
... and 13 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 76abfa4...f04dc13. Read the comment docs.

[pep8speaks]

Hello @glemaitre! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

• In the file examples/ensemble/plot_bagging_classifier.py:

Line 28:1: E402 module level import not at top of file
Line 39:1: E402 module level import not at top of file
Line 50:1: E402 module level import not at top of file
Line 51:1: E402 module level import not at top of file
Line 69:1: E402 module level import not at top of file
Line 70:1: E402 module level import not at top of file
Line 79:1: E402 module level import not at top of file
Line 97:1: E402 module level import not at top of file
Line 121:1: E402 module level import not at top of file
Line 122:1: E402 module level import not at top of file
Line 123:1: E402 module level import not at top of file

• In the file examples/ensemble/plot_comparison_ensemble_classifier.py:

Line 31:1: E402 module level import not at top of file
Line 32:1: E402 module level import not at top of file
Line 50:1: E402 module level import not at top of file
Line 57:1: E402 module level import not at top of file
Line 58:1: E402 module level import not at top of file
Line 67:1: E402 module level import not at top of file
Line 68:1: E402 module level import not at top of file
Line 85:1: E402 module level import not at top of file
Line 86:1: E402 module level import not at top of file
Line 114:1: E402 module level import not at top of file
Line 134:1: E402 module level import not at top of file
Line 135:1: E402 module level import not at top of file
Line 182:1: E402 module level import not at top of file
Line 183:1: E402 module level import not at top of file

• In the file imblearn/ensemble/_bagging.py:

Line 251:13: W503 line break before binary operator

Comment last updated at 2021-02-18 11:24:39 UTC