# Get the prediction probability using prediction function

I'm new to SVM models. I took custom SVM classifier from the github. In there standard predict function was overwritten by custom predict function.

def predict(self, instances):
predictions = []
for instance in instances:
predictions.append((int)(np.sign(self.weight_vector.dot(instance.T)[0] + self.bias)))#class is determined based on the sign -1,1


I want to get prediction probability of each class. [-1 probability, 1 probability] I know from standard SVM predict_proba function I can get the probability. Since this is a customized SVM how do I get that? Can somebody help me.

Thank you