CODEWITHSUNDEEP
pythonroc
Alan
Alan

Reputation: 129

KeyError when plotting ROC curve

I am trying to plot the ROC curve for a multi label classification problem. My target column before giving it labels looked like:

TargetGrouped
I5
I2
R0
I3

This is part of the code used to plot the ROC:

    def computeROC(self, n_classes, y_test, y_score):
        # Compute ROC curve and ROC area for each class
        fpr = dict()
        tpr = dict()
        roc_auc = dict()

        # Compute micro-average ROC curve and ROC area
        fpr["micro"], tpr["micro"], _ = roc_curve(y_test.ravel(), y_score.ravel())
        roc_auc["micro"] = auc(fpr["micro"], tpr["micro"])

        # Plot ROC curve
        plt.figure()
        plt.plot(fpr["micro"], tpr["micro"],
                 label='micro-average ROC curve (area = {0:0.2f})'
                       ''.format(roc_auc["micro"]))
        for i in range(n_classes):
            plt.plot(fpr[i], tpr[i], label='ROC curve of class {0} (area = {1:0.2f})'
                                           ''.format(i, roc_auc[i]))

        plt.plot([0, 1], [0, 1], 'k--')
        plt.xlim([0.0, 1.0])
        plt.ylim([0.0, 1.05])
        plt.xlabel('False Positive Rate')
        plt.ylabel('True Positive Rate')
        plt.title('Some extension of Receiver operating characteristic to multi-class')
        plt.legend(loc="lower right")
        plt.show()

    lb = preprocessing.LabelBinarizer()
    infoDF = infoDF.join(pd.DataFrame(lb.fit_transform(infoDF["TargetGrouped"]), columns = lb.classes_, index = infoDF.index))

#Extracted features and split infoDF dataframe => got X_train, X_test, y_train, y_test

    rfc = RandomForestClassifier()
    rfc.fit(X_train, y_train)
    y_pred = rfc.predict(X_test)

    classes = y_train.shape[1]
    computeROC(classes, y_test, y_pred)

When I run it, I get the following error:

Traceback (most recent call last):

  File "<ipython-input-50-15a83ece5e44>", line 3, in <module>
    evaluation.computeROC(classes, y_test, y_pred)

  File "<ipython-input-49-526a19a07850>", line 18, in computeROC
    plt.plot(fpr[i], tpr[i], label='ROC curve of class {0} (area = {1:0.2f})'

KeyError: 0

I know it might be an issue with the format of the class in the plt.plot() but I don't know how to fix it.

Update

This is the code missing from the computeROC():

        for i in range(n_classes):
            fpr[i], tpr[i], _ = roc_curve(y_test[:, i], y_score[:, i])
            roc_auc[i] = auc(fpr[i], tpr[i])

Upvotes: 0

Views: 1079

Answers (1)

Gianluca Micchi
Gianluca Micchi

Reputation: 1653

The issue is actually with the following line:

for i in range(n_classes):
            plt.plot(fpr[i], tpr[i], label='ROC curve of class {0} (area = {1:0.2f})'
                                           ''.format(i, roc_auc[i]))

fpr and tpr are dictionaries and the only key you initialized is 'micro'. This for loop assigns integer values between 0 and n_classes-1 to i, but you have never defined what fpr[0] and tpr[0] are (and I suspect you are thinking of them as lists, but this is just speculation).

Upvotes: 1

Related Questions