'sample_weight' error using Adaboost Classifier with logistic regression and Naive Bayes

[IssamLaradji]

I'm peculiarly getting an error while using AdaBoostClassifier with GaussianNB() or LogisticRegression() as a base estimator.

These are my commands

In [65]: gnb = GaussianNB()
In [66]: bdt = AdaBoostClassifier(gnb,n_estimators=100)
In [67]: bdt.fit(X,y)

I get the following error after executing "In[67]" :

TypeError: fit() got an unexpected keyword argument 'sample_weight'

Any reason why I might be getting this?

PS: I frequently use adaboosting with Navie Bayes as a base classifier in WEKA, hence the concern

Thank you very much!

[glouppe]

Hello,

This is not a bug.

Unfortunately, you cannot use any base estimator with AdaBoostClassifier. Only those that support a "sample_weight" fit argument are allowed (e.g., DecisionTreeClassifier or SGDClassifier).

I am closing this non-issue.

[NelleV]

It this is a non issue, it should at least be mentionned in the documentation on which classifier this can be used.

[NelleV]

(And there should be a proper error message).

[glouppe]

@NelleV It is already mentionned in the documentation. See base_estimator at http://scikit-learn.org/dev/modules/generated/sklearn.ensemble.AdaBoostClassifier.html

[NelleV]

@glouppe I've read the documentation, and it doesn't seem obvious to me which classifier I can and cannot use with this (I actually digged in the sourcecode to find out). I think we should list explicitely what can be used here.

[cayoglu]

What a nice Idea :) What happened?

[jnothman]

I see "support for sample weighting is required" in the class documentation

…

On 14 Mar 2017 3:07 pm, "cayoglu" ***@***.***> wrote: What a nice Idea :) What happened? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1752 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAEz66jZYzPTDZSvLU8UalR6JloXfxZ3ks5rlhJtgaJpZM4AfAvs> .

[agrija9]

There is this answer in Stackoverlow that lists the supported base estimators for Adaboost.

If you're in Python 3.6 just replace the "clf().fit" with "clf.fit".