Most of the data ingestion utilities (pd.read_csv, pd.read_parquet, pd.read_sql) since Pandas 2.2 include a dtype_backend='pyarrow' option to use PyArrow types instead of NumPy backend.
It would be awesome if there was support for this.
Hey @mattharrison , thank you for raising this issues and my apologies for the delay in response. I will discuss this request with the team and provide an update here.
Most of the data ingestion utilities (
pd.read_csv
,pd.read_parquet
,pd.read_sql
) since Pandas 2.2 include adtype_backend='pyarrow'
option to use PyArrow types instead of NumPy backend.It would be awesome if there was support for this.