can someone help me fix it ?

Hi @Francisco can you tell use what version of StatsForecast you're using? Can you also check the format of train_agg and forecast_df? They should be like in this example

yeah everything with the data and the format is good except for the statsforecast library, it seems it isnt importing well

I see. Can you check if you're working with version 1.6.0 of StatsForecast?

pip show statsforecast

I used an AutoARIMA with a subset of the M4 Hourly dataset and was able to generate the plots below.

It's a matplotlib error, you probably have a broken environment