add MAD outlier rule to find outliers

[glevv]

PR for #496
Added MAD outlier rule
Refactored base code and tests

Extend outlier detection functionality of OutlierTrimmer and Winsorizer.

[solegalli]

Hi @glevv

Really, really good implementation of the outlier handlers. Thank you very much for taking the time to re-write almost the entire logic. Very elegant. I appreciate it a lot!

I also learned a lot.

There are a couple of tests failing.

Stylechecks is complaining about numpy imported but being unused, see below:

stylechecks run-test: commands[0] | flake8 feature_engine tests
feature_engine/outliers/base_outlier.py:3:1: F401 'numpy as np' imported but unused
feature_engine/outliers/base_outlier.py:179:89: E501 line too long (89 > 88 characters)
feature_engine/outliers/base_outlier.py:195:89: E501 line too long (91 > 88 characters)
feature_engine/outliers/base_outlier.py:218:89: E501 line too long (90 > 88 characters)
feature_engine/outliers/trimmer.py:4:1: F401 'numpy as np' imported but unused
tests/test_outliers/test_outlier_trimmer.py:4:1: F401 'numpy as np' imported but unused

The Python tests are failing because of the error to catch when scale

This test is also failing:

feature_engine/tests/test_outliers/test_winsorizer.py

Line 303 in 65fedf1

def test_error_if_capping_method_quantiles_and_fold_value_not_permitted():

I believe they should all be small fixes. Would you be able to have a look at it?

I think once the tests pass this is ready to go!

Thank you very much!

[glevv]

Fixes implemented

[solegalli]

Hey @glevv

This implementation is great. Thank you so much!

Code-wise, this is good to go.

If I may ask a little bit more from you, could you please add the Mad rule to the user guide? In this 2 files:

https://github.com/feature-engine/feature_engine/blob/main/docs/user_guide/outliers/Winsorizer.rst

https://github.com/feature-engine/feature_engine/blob/main/docs/user_guide/outliers/OutlierTrimmer.rst

Thank you!