search

Ritam-Guha / Py_FS

A Python Package for Feature Selection
MIT License
57 stars 38 forks source link

Regarding Filter method #30

Closed akshitamishra1213 closed 1 year ago

akshitamishra1213 commented 2 years ago

Hello There, I am a bit new to all these python things but thanks to your wonderful code I am getting a bit of confidence. I tried to use Filter method Pearson Correlation Coefficient. Got feature ranks and all. But I couldn't understand how to prepare my dataset with the selected best features. If you could please guide.

akshitamishra1213 commented 2 years ago

or can you give a demo of how to read a dataset use one of these filters and create improved dataset plzzz

Ritam-Guha commented 2 years ago

So, we have an API to do that. If you go to the Google Colab page for the demonstration: https://colab.research.google.com/drive/1PafNTmVgWv9Qz6j7bI41XqPT6CCCIb1T?usp=sharing#scrollTo=RJhTgmqhienI, You will find that we are printing something called solution.ranked_features for the filter part. That basically represents your features in ranked order. That means the first column for that matrix represents the best feature according to our process. If you want to use like 10 features, then you can just get the features like solution.ranked_features[:, :10]. Let me know if you have any doubt about this. ☺️

akshitamishra1213 commented 2 years ago

Thank you for Quick response. :)