I have been using the scikit-learn library. I'm trying to use the Gaussian Naive Bayes Module under the scikit-learn library but I'm running into the following error. TypeError: cannot perform reduce with flexible type
Below is the code snippet.
training = GaussianNB()
training = training.fit(trainData, target)
prediction = training.predict(testData)
This is target
['ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'ALL', 'AML', 'AML', 'AML', 'AML', 'AML', 'AML', 'AML', 'AML', 'AML', 'AML', 'AML']
This is trainData
[['-214' '-153' '-58' ..., '36' '191' '-37']
['-139' '-73' '-1' ..., '11' '76' '-14']
['-76' '-49' '-307' ..., '41' '228' '-41']
...,
['-32' '-49' '49' ..., '-26' '133' '-32']
['-124' '-79' '-37' ..., '39' '298' '-3']
['-135' '-186' '-70' ..., '-12' '790' '-10']]
Below is the stack trace
Traceback (most recent call last):
File "prediction.py", line 90, in <module>
gaussianNaiveBayes()
File "prediction.py", line 76, in gaussianNaiveBayes
training = training.fit(trainData, target)
File "/Library/Python/2.7/site-packages/sklearn/naive_bayes.py", line 163, in fit
self.theta_[i, :] = np.mean(Xi, axis=0)
File "/System/Library/Frameworks/ core/fromnumeric.py", line 2716, in mean
out=out, keepdims=keepdims)
File "/System/Library/Frameworks/", line 62, in _mean
ret = um.add.reduce(arr, axis=axis, dtype=dtype, out=out, keepdims=keepdims)
TypeError: cannot perform reduce with flexible type
2 Answers
It looks like your 'trainData' is a list of strings:
['-214' '-153' '-58' ..., '36' '191' '-37']
Change your 'trainData' to a numeric type.
import numpy as np
np.array(['1','2','3']).astype(np.float)
When your are trying to apply prod on string type of value like:
['-214' '-153' '-58' ..., '36' '191' '-37']
you will get the error.
Solution:
Append only integer value like [1,2,3], and you will get your expected output.
If the value is in string format before appending then, in the array you can convert the type into int type and store it in a list.
