Open
Description
Overview: Add functionality to get column types, including date, and sortkeys from Redshift meta info for Pandas IO
I modified the client to allow me to pass arguments to the pd.read_csv on line 227 in civis/io/_tables.py. I gather the column types using client.tables.get after which I map the SQL types to their corresponding Python types. Doing this eliminates the automatic str to int conversion, e.g. zipcodes, and automatically identifies the date columns. Is there any interest in adding this in a more thoughtful way to the code?