How to add sklearn shell for new ML algorithm

Question

I would like to integrate factorization machines into sklearn. I checked the sklearn documentation and the website on how to wrap the new algorithm, but this requirement does not seem to be very well documented.

So, I would like to ask if there is documentation on how to add a new algorithm shell to sklearn (besides reading the source code)?

Answer 1

After working with the sklearn documentation, sklearn best to look at the full working example.

The XGBoost module has an extensive sklearn shell, which you can see here:

https://github.com/dmlc/xgboost/blob/master/python-package/xgboost/sklearn.py

Answer 2

From this FAQ, I understand that they do not really like the new http://scikit-learn.org/stable/faq.html#selectiveness algorithms for reasons that seem valid. Given this, it is plausible to think that there is no documentation on how to add a new algorithm shell. I will add that I have been using this package for a long time, and I have never found anything on my website or on other websites that looked like what you are looking for.