TypeError: 'numpy.float64' object is not callable - While Printing F1 Score

Question

I am trying to run below code on Jupyter Notebook:

lr = LogisticRegression(class_weight='balanced')
lr.fit(X_train,y_train)
y_pred = lr.predict(X_train)

acc_log = round(lr.score(X_train, y_train) * 100, 2)
prec_log = round(precision_score(y_train,y_pred) * 100,2)
recall_log = round(recall_score(y_train,y_pred) * 100,2)
f1_log = round(f1_score(y_train,y_pred) * 100,2)
roc_auc_log = roc_auc_score(y_train,y_pred)

When trying to execute this, I am getting the below error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-46-bcb2d9729eb6> in <module>
      6 prec_log = round(precision_score(y_train,y_pred) * 100,2)
      7 recall_log = round(recall_score(y_train,y_pred) * 100,2)
----> 8 f1_log = round(f1_score(y_train,y_pred) * 100,2)
      9 roc_auc_log = roc_auc_score(y_train,y_pred)

TypeError: 'numpy.float64' object is not callable

Can't seem to figure out what I am doing wrong.

Wilinski

Answer 1

Somewhere in your code (not shown here), there is a line which says f1_score = ... (with the written type being numpy.float64) so you're overriding the method f1_score with a variable f1_score (which is not callable, hence the error message). Rename one of the two to resolve the error.

Touchhole

That seemed to be exactly what the problem was. Removed that code and restarted the kernel and ran again. Worked. – Wilinski

Answer 2

Use metrics.f1_score(y_train,y_pred) instead of f1_score(y_train,y_pred) in this situation:

demo screenshot

Seibert

Which will bypass the problem exposed in @runDOSrun's answer. It's a way to fix the problem, indeed, but I would personaly change the variable name. If you won't, you'll probably face the same problem again later – Membrane