CODEWITHSUNDEEP
pythonscikit-learndata-analysis
MLhacker
MLhacker

Reputation: 1512

Fitting Naive Bayes on Floating Point Data Values

I want to fit Gaussian Naive Bayes on data values with floating point, and the code I'm using is this:

array1 = np.array([[2,2],[3,2]]) 
array2 = np.array([0.3,3])

clf = GaussianNB()
clf.fit(array1,array2)

But, I get an error saying:

ValueError("Unknown label type: %s" % repr(ys)) ValueError: Unknown label type: (array([ 0.3, 3. ]),)

How can I get around the issue without using a different naive bayes module than the one provided by Sklearn?

Upvotes: 0

Views: 1452

Answers (1)

Farseer
Farseer

Reputation: 4172

You using array2 as your target labels.

GaussianNB() is a classifier, so target labels must be integers.(in your case 0.3 is float)

If your labels are real numbers, consider using Regression.

Upvotes: 2

Related Questions