Migrate pickled dataframes to pandas >=2.0

[millsks]

Thank you so much for providing this dataset. It proved to be very useful in a ML POC that we were working on using Kedro. One issue that we noticed was the pickled dataframes were created using a pre-2.0 version of pandas which causes the following error to occur when the files is read. Since the pandas 2.0 release broke pandas 1.x backwards compatability the internal pandas.core.indexes.numeric no longer existed. It would be beneficial to those who are using the dataset to have the files pickled using the latest supported version of pandas.

DatasetError: Failed while loading data from data set PickleDataSet(backend=pickle,
filepath=/path/to/kedro/project/data/01_raw/2018-04-01.pkl, load_args={}, protocol=file, save_args={}).   
No module named 'pandas.core.indexes.numeric'