ModuleNotFoundError: No module named 'skfeature'

Question:

I am unable to import fisher score from skfeature.function.similiarity_based.

Is there a problem in my code below?

from skfeature.function.similarity_based import fisher_score
ranks = fisher_score(X,Y)
feature_importance = pd.Series(ranks,df.columns[0:len(df.columns) - 1])
feature_importance.plot(kind = 'barh', color = 'magenta')
plt.show()

I’m working on Kaggle notebook not locally.

I’ve tried installing it using pip commands and have tried to look for other alternatives but nothing seems to work.

Asked By: Prayas Gautam

||

Answers:

That’s a third-party library not associated with scikit-learn.

Someone uploaded a version to PyPi (PyPi, GitHub):

pip install skfeature-chappers

Or the original should be possible to install directly from GitHub with:

pip install git+https://github.com/jundongl/scikit-feature.git
Answered By: Alexander L. Hayes