How can I resolve this Error - " ValueError: Negative values in data passed to MultinomialNB (input X)

Question

model_3 = MultinomialNB()
model_3.fit(X_train,np.ravel(y_train))
y_predict = model_3.predict(X_test)
accuracy = metrics.accuracy_score(y_test,y_predict)
print(accuracy)

I am getting an error:

ValueError: Negative values in data passed to MultinomialNB (input X)

How can I resolve this error ?

You are giving model_3 a negative number, which is not allowed. — Libra, Apr 23 '21 at 05:09
Would help if you could check (eg by printing) what value you are passing to MultinomialNB. — Mister Mak, Apr 23 '21 at 07:30

score 2 · Answer 1 · edited Mar 18 '22 at 12:01

2

Try using MinMaxScaler() to preprocess the data before sending the data to the model. This normalizes it to the range 0 to 1 thus removing the negative numbers.

from sklearn.preprocessing import MinMaxScaler #fixed import

scaler = MinMaxScaler()
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)

edited Mar 18 '22 at 12:01

Community

1
1

answered Apr 23 '21 at 05:11

yudhiesh

6,383
3
16
49

How can I resolve this Error - " ValueError: Negative values in data passed to MultinomialNB (input X)

1 Answers1