https://github.com/nixtla logo
#general
Title
# general
d

Daniel Śliwiński

10/11/2023, 1:56 PM
Hi guys, I'm trying to train AutoARIMA model with exogenous variables using spark in databricks. I have added exogenous variables after the y variable as suggested. train_df_monthly is in format of Spark dataframe. What options do I have, if I want to use spark dataframe and not pandas, as my dataset is really big. Kind regards, Daniel I am executing this code: # generate forecasts sf = StatsForecast( models = [AutoARIMA(season_length=season_length)], freq = 'MS', n_jobs=-1, fallback_model=Naive() ) y_pred = sf.forecast( df = train_df_monthly, h = h ) And receiving this error: PythonException: 'KeyError: "['EX_VAR1', 'EX_VAR2'] not in index"'. Full traceback below:
j

José Morales

10/11/2023, 3:48 PM
Hey. This was an issue and has been fixed in the main branch but hasn't made it to a release yet, are you able to install from github?
2 Views