I am trying to use the statsmodels Python library, and in particular the ARIMA model. I have some troubles performing out-of-sample forecast with it. Here is my call:
predicted = my_arima_result.predict(start=startDateTime, end=endDateTime)
From my understanding, I need to pass a start and end parameters; but when passing a end parameter I get the following error: ValueError: no rule for interpreting end
I did some tests and the call only works for in-sample calls, passing at most the start parameter.
The model is fit with float dataframe, indexed by a datetime, at a frequency of 6 minutes.
I am using version 0.8.0rc1 of statsmodels installed through easyinstall, which seems to be the latest version... but I cannot find the last code line at the origin of the Error in the documentation/git...
