CODEWITHSUNDEEP
pythonmachine-learningscikit-learnclassificationfeature-selection
Anas El Kesri
Anas El Kesri

Reputation: 51

Troubles using SelectKBest [scikit-learn]

i'm a beginner with scikit-learn and python, i tried this code that looks very simple using SelectKBest from feature_selection package.

    train = pd.read_csv("train.csv")
test = pd.read_csv('test.csv')
train = SelectKBest(chi2, k=120).fit_transform(train)

print train.shape

but i got this error and i really don't know how to fix it !!

TypeError: fit() takes exactly 3 arguments (2 given)

Can u guys please help me fixing that :( ?

Upvotes: 2

Views: 758

Answers (1)

Oer
Oer

Reputation: 447

The issue is that it expect 3 arguments.

fit(X, y) X : array-like, shape = [n_samples, n_features] which is the training input samples. y : array-like, shape = [n_samples] is the target values (class labels in classification, real numbers in regression).

fit_transform(X,y=None, **fit_params)[source]

Therefore, to fix your problem, if the labels are stored in the last column of the training/test data i.e. train.ix[:,-1], you could do this:

train = pd.read_csv("train.csv")
test = pd.read_csv('test.csv')
model = SelectKBest(chi2, k=120).fit_transform(train.ix[:,:-1],train.ix[:,-1])

print train.shape

This code worked for me.

Upvotes: 1

Related Questions