Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

CategoricalEncoder in preprocessing or feature_extraction module?

See original GitHub issue

In the merged PR, the CategoricalEncoder was put in the preprocessing module. However, it might make more sense to put it in feature_extraction since it is dealing with extracting numerical features from categorical data (eg also DictVectorizer is in that module). I think the main reason it is in preprocessing is because OneHotEncoder is historically there.

But since it’s not yet released we can still change this, if we would want to.

cc @amueller @jnothman @GaelVaroquaux

Issue Analytics

State:
Created 6 years ago
Comments:6 (6 by maintainers)

Top GitHub Comments

2reactions

jnothmancommented, May 23, 2018

I think this can be closed, regardless of whether there is clear consensus.

Let me join your killing spree…

1reaction

jorisvandenbosschecommented, May 24, 2018

It is of course not exact by date, but you always look at https://github.com/scikit-learn/scikit-learn/pulse. Anyhow, the balance for this week is -40 issues 😃

Top Results From Across the Web

sklearn.preprocessing.OneHotEncoder

Encode categorical features as a one-hot numeric array. ... By default, the encoder derives the categories based on the unique values in each...

feature_engine Documentation - Read the Docs

Feature-engine is a Python library with multiple transformers to engineer and select features to use in machine learning models.

Categorical Encoder in Scikit Learn Preprocessing

I have version 0.21.3 of sklearn. I checked online to see the documentation and it seems that CategoricalEncoder was there in version 0.20.dev0 ......

scikit-learn : Data Preprocessing I - Missing/categorical data

scikit-learn : Data Preprocessing (missing/categorical data) ... scikit-learn : Features and feature extraction - iris dataset

Feature-engine — 1.5.2

Variable selection. Datetime features. Time series. Preprocessing. Feature-engine allows you to select the variables you want to transform ...