At the moment, we only use pandas to handle dataframes.
It could become handy (especially with big data frames) to have different backends available for handling and processing dataframes.
These backends could be pandas, polars and dask.
Replacing pandas is especially important in the io extension, but there are many more places, where pandas is used.
[ ] implement backend functionality for switching dataframe classes
At the moment, we only use
pandasto handle dataframes.It could become handy (especially with big data frames) to have different backends available for handling and processing dataframes.
These backends could be
pandas,polarsanddask.Replacing pandas is especially important in the io extension, but there are many more places, where pandas is used.
Entry.get_data()uses dataframe backendImmutableResultSet.get_data()uses dataframe backend