-
#### Describe the bug
I'm trying to load a fitted SARIMAX model I saved earlier to a .pkl file and I get the following error:
AttributeError: Can't get attribute 'FitStopper' on
I'm using Spyd…
-
see: https://github.com/statsmodels/statsmodels/issues/6225#issuecomment-550350284
-
in this commit: [https://github.com/statsmodels/statsmodels/commit/d03474da1aae5dac54c2b4441311d01517cd2567](url) the sarimax model was changed to only warn on non-invertible/stationary start_params a…
-
#### Is your feature request related to a problem? Please describe
I'm trying to re-write SAS code into a python script. SAS allows for me to select specific lags (ex, p = (1,3,6)) vs. lag order (ex,…
-
Requested via e-mail. It's exposed now as `_arima_predict_out_of_sample`, it's just a private function and not documented.
-
Hello everyone!
May be its not a bug, but, I don't undirstand how you compute starting points in AR model.
For example:
I want to estimate ar model like:
y=AR(1)+AR(2)+e
get coeficients, fitted …
-
Hi team i am trying to fit a time series model with seasonal length 365 days and notice it is taking actually over 2 hours to run. The model parameters look like this and historical_dat is a data tabl…
-
at sarimax.py, line 953:
`If we have estimated non-stationary start parameters but enforce stationarity is on, raise an error` is misleading. Actually the code just warn and then zero out the coeffi…
-
#### Describe the bug
SARIMAX does not work with directly specifying lags. Appears to have a bug in starting value code.
This is using latest dev (Nov 12, 2019)
#### Code Sample, a copy-pasta…
-
Training on recordings of motion tracking for mice showed occasional momentary deviations from the true mouse position. This can probably be fixed in post processing by identifying outlying data point…