nukepanda
nukepanda

Reputation: 1

How to get fitted in-sample values using auto_arima in python?

I have no problem in getting the forecast out-of sample values but I can't seem to find a way to show the fitted in-sample ones.

model = auto_arima(y = training_set['Y'],
                  m = 12,
                  seasonal = True,
                  stepwise = False)

predictions = model.predict(n_periods = test_period)
predictions

Upvotes: 0

Views: 1142

Answers (1)

Draco
Draco

Reputation: 41

You can use the predict_in_sample() method of your ARIMA model object

samples = model.predict_in_sample()

You can also retrieve the in sample values between 2 integer positions

predict_in_sample: https://alkaline-ml.com/pmdarima/modules/generated/pmdarima.arima.ARIMA.html?#pmdarima.arima.ARIMA.predict_in_sample

It appears there was an issue with this model that has been resolved in the current version. https://github.com/alkaline-ml/pmdarima/issues/140

Upvotes: 0

Related Questions