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.

Feature Discussion: FeatureUnion with axis choosing

See original GitHub issue

Hi, I am using FeatureUnion for a transformation Pipeline. FeatureUnion is always using np.hstack to concatenate the results. See here: https://github.com/scikit-learn/scikit-learn/blob/bac89c2/sklearn/pipeline.py#L829

I would like to discuss the following idear: What about using numpy.concatenate and let the user choose the axis with a maybe optional parameter? What do you think about this? I can do a PR but wanted to discuss this before I start coding.

Thanks Philip

Issue Analytics

State:
Created 5 years ago
Comments:10 (3 by maintainers)

Top GitHub Comments

2reactions

amuellercommented, Nov 21, 2018

pipegraph and nimbusml both are more flexible.

But there’s no reason why you can’t use scikit-learn’s pipeline for this with your own transformer. Looks like you implemented it already. The pipeline is completely agnostic in terms of dimensionality of the data. The issue is more with cross-validation. If you think it’s more widely useful you can publish your implementation yourself, like many other estimators in the scikit-learn-contrib org.

0reactions

PhilipMaycommented, Oct 2, 2020

Well - closing it…

Top Results From Across the Web

sklearn.pipeline.FeatureUnion

This estimator applies a list of transformer objects in parallel to the input data, then concatenates the results. This is useful to combine...

Scikit-learn utility function to select specific columns in a pipeline

The ColumnSelector can be used for "manual" feature selection, e.g., as part of a grid search via a scikit-learn pipeline.

Feature selection (MDI, PERM, RFE) in depth review - Kaggle

In this notebook, we are going to look at tree-based impurity reduction feature importance, permutation feature importance and reccurent feature elimination.

3 Healthcare: Diagnosing COVID-19 - liveBook · Manning

Analyzing tabular data to judge which feature engineering techniques are going to help; Implementing feature improvement, construction, and selection ...

Time-related feature engineering with scikit-learn - Neuraxle

... using the Pipeline class, as well as ColumnTransformer and FeatureUnion . ... on the test set in the discussion, we instead choose...