I tried both Python 2.7.14 and Python 3.6.4. pandas==0.22.0 and xlrd==1.1.0. https://transfer.sh/2T4fw/hourly_loads.csv.gz https://transfer.sh/2T4fw/hourly_load

I tried both Python 2.7.14 and Python 3.6.4. pandas==0.22.0 and xlrd==1.1.0. https://transfer.sh/2T4fw/hourly_loads.csv.gz https://transfer.sh/2T4fw/hourly_loads.csv.gz is what I get. It looks like dates are stored not as dates but strings in the more recent spreadsheets. edit: dateutil's parser.parse(x) & strftime("%Y-%m-%d") helps