Dominick dataset forecasting data, weekly data.
From original source:
-----
This dataset contains 115704 weekly time series representing the profit of individual stock keeping units from a retailer.
-----
It is not clear which of the 'movement' files from https://www.chicagobooth.edu/research/kilts/research-data/dominicks# was used.
They also did not keep a track of the date.
There are 3 columns:
id_series: The identifier of a time series.
value_0: The value of the time series at 'time_step'.
time_step: The time step on the time series.
Preprocessing:
1 - Renamed columns 'series_name' and 'series_value' to 'id_series' and 'value_0'.
2 - Exploded the 'value_0' column.
3 - Created 'time_step' column from the exploded data.
4 - Defined 'id_series' as 'category' and casted 'value_0' to float.