feature_engine

feature_engine copied to clipboard

add function to check that y is binary in dataframe_checks

2

Many transformers in feature engine require that y is binary. At the moment we do this check within each transformer. We should create a function that unifies this behaviour and...

solegalli

include code coverage

2

At the moment, we have no visibility of test coverage. I would like to include this in the package. Here some useful links I gathered: https://gist.github.com/dnozay/5b9b818ff0dc857d1358 https://pytest-cov.readthedocs.io/en/latest/tox.html https://coverage.readthedocs.io/en/6.1.1/config.html Also check...

solegalli

Information value for feature selection using weight of evidence (WoE)

19

Information value is a variable-selection method that is used with binary classifiers. Information value summarizes how much knowing `var_A` helps in predicting the dependent variable. feature-engine includes a WoE Encoder....

Morgan-Sell

add missing_only=True to all imputers to use in combination with variables=None

13

Add missing_only functionality to all imputers to use in combination with variables=None When variables is None, the imputers select all numerical, or categorical or all variables by default. With the...

solegalli

new discretization transformer: IRD or OneRuleDiscretizer

1

Holte 1993. More details coming soon

solegalli

[Feature Creation] Decision tree creates a new feature by combining numerous variables

16

Closes #107 Notes from #107: New variables are created by combination of user indicated variables with decision trees. Example: if user passes 3 variables to transformer, a new feature will...

Morgan-Sell

[New Transformer] ChiMergeDiscretisor

7

Closes #450. Notes from #450: Existing implementations: https://github.com/lisette-espin/pychimerge https://github.com/night18/ChiMerge https://github.com/raiyan1102006/ChiMerge https://gist.github.com/alanzchen/17d0c4a45d59b79052b1cd07f531689e?short_path=f2e54c6 Reference to the original article can be found in the first link

Morgan-Sell

Handling of missing values in categorical encoders

6

Some transformers such as the `CountFrequency` categorical encoder have options for handling new unseen categories during calls to `transform`. In some cases, unseen categories and explicit missing values might be...

david-cortes

(In progress) Backoff binning rare labels in CountFrequency encoder

6

**Reference Issues/PRs**: Issue : #429 (Add functionality for threshold) **What does this implement/fix?** Adds functionality to group categories with few observations. Hi, @solegalli I've added functionality for Backoff binning, threshold...

SangamSwadiK

new discretiser: sorts variables in bins, and replaces bins by target mean value

4

We can use the functionality of the target mean classifiers or regressor probably.

solegalli