In doing so broke the BPA client which pases keyword arguments into BaseClient.parse_to_df(...) to read Excel files. I took a stab with #174 but the proper fix is to actually:
Change requirements.txt and setup.py to pandas>=0.21 when we're ready.
Fix the BaseClient, BPAClient, and BPA unit tests to use usecols rather than parse_cols.
Run both unit tests and integration tests for the BPA client to ensure things are working properly again.
Pandas 0.21 came out on October 27th and deprecated
read_excel(parse_cols=[])
in favour ofread_excel(usecols=[])
.https://pandas.pydata.org/pandas-docs/version/0.21/whatsnew.html#deprecations
In doing so broke the BPA client which pases keyword arguments into
BaseClient.parse_to_df(...)
to read Excel files. I took a stab with #174 but the proper fix is to actually:pandas>=0.21
when we're ready.BaseClient
,BPAClient
, and BPA unit tests to useusecols
rather thanparse_cols
.