I am dealing with time series and using statsmodel's ARIMA to fit a model using maximum likelihood estimation. When I call ARIMAResult.summary(), I see the estimated parameters, one of them being the variance of the error term.

Is there a way to get a probability density for each observation in my time series, assuming that it is generated by my estimated model?

I have looked through the documentation and have only found functions that give me the likelihood of my parameters but could not find anything for the probability of the series. I see that I can calculate what I want by using ARIMAResult.resid and the estimated variance. But why is there no built-in function for this?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Browse other questions tagged or ask your own question.