UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples. 'recall', 'true', average, warn_for)

Question

When I use the following code to calculate precision_recall_fscore_support for one-class ( only the 1s)

import numpy as np
from sklearn.metrics import precision_recall_fscore_support

#make arrays
ytrue = np.array(['1', '1', '1', '1', '1','1','1','1'])
ypred = np.array(['0', '0', '0', '1', '1','1','1','1'])

#keep only 1
y_true, y_pred = zip(*[[ytrue[i], ypred[i]] for i in range(len(ytrue)) if ytrue[i]=="1"])

#get scores
precision_recall_fscore_support(y_true, y_pred, average='weighted')

I get the following Warning:

UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples.
  'recall', 'true', average, warn_for)

and output:

(1.0, 0.625, 0.76923076923076927, None)

I found the following on SO UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples that has similar warning but I don't think it applies to my problem.

Question: Are the results of my output valid or should I be concerned about the warning message? If so, what is wrong with my code and how to fix?

Read how these metrics are define and it will be clear why you can't compute them with your current data. It's like trying to compute the proportion of female smokers out of a male only sample. Makes no sense. — Julien, Feb 26 '18 at 01:41
I tried the following code below. However, it was downvoted but I am not sure why. — , Feb 26 '18 at 20:08

score 0 · Accepted Answer · edited Dec 14 '18 at 04:26

hello i found the solution this problem, you need use:

cv = ShuffleSplit(n_splits=10, test_size=0.3, random_state=0)

I'm using knn and this solved the problem

Code:

def knn(self,X_train,X_test,Y_train,Y_test):

   #implementación del algoritmo
   knn = KNeighborsClassifier(n_neighbors=3).fit(X_train,Y_train)
   #10XV
   cv = ShuffleSplit(n_splits=10, test_size=0.3, random_state=0)
   puntajes = sum(cross_val_score(knn, X_test, Y_test, 
                                        cv=cv,scoring='f1_weighted'))/10

   print(puntajes)

**Link: ** https://scikit-learn.org/stable/modules/cross_validation.html

UndefinedMetricWarning: Recall and F-score are ill-defined and being set to 0.0 in labels with no true samples. 'recall', 'true', average, warn_for)

1 Answers1