dmbee / seglearn

Python module for machine learning time series:
https://dmbee.github.io/seglearn/
BSD 3-Clause "New" or "Revised" License
571 stars 63 forks source link

FeatureRepMix #11

Closed qtux closed 5 years ago

qtux commented 5 years ago

Hi David,

I implemented a FeatureRepMix as discussed in #9 supporting contextual data. Instead of deriving from ColumnTransformer, I derived the class from _BaseComposition in order to have a less hacky solution (no need to disable unwanted features as in the previous solution) and combined it with large parts of the FeatureRep implementation. The only missing feature is the parallel execution of all FeatureRep transformers which could easily be added in the future.

Likewise, I changed the example and added tests in close resemblance to the FeatureRep tests. I hope that you are fine with that, as I am currently a bit short of time for more sophisticated tests.

Cheers, Matthias

coveralls commented 5 years ago

Pull Request Test Coverage Report for Build 136


Changes Missing Coverage Covered Lines Changed/Added Lines %
seglearn/transform.py 53 61 86.89%
<!-- Total: 81 89 91.01% -->
Totals Coverage Status
Change from base Build 133: -0.2%
Covered Lines: 1396
Relevant Lines: 1484

💛 - Coveralls
dmbee commented 5 years ago

This is great Matthias. Nicely done.

Can you send me a suitable signature block for this: https://dmbee.github.io/seglearn/about.html so I can add you as a contributor.

Thanks