I am comparing a model's outcome with the actual values for a span of 9 months. I am expecting a very large error, however my MAPE implementation yields 1404512.56 using python, pandas.
Here is my implementation:
def mean_absolute_percentage_error(y_true, y_pred):
y_true, y_pred = np.array(y_true), np.array(y_pred)
return np.mean(np.abs((y_true - y_pred) / y_true)) * 100
There were instances in y_true (the actual values) where they were 0, but I have replaced them with 0.001 to avoid division by 0.
If anyone could provide a hint as to what MAPE means when it's that high, I'd appreciate it.
