drop_duplicates not working in pandas?

I have just had this issue, and this was not the solution.

It may be in the docs - I admittedly havent looked - and crucially this is only when dealing with date-based unique rows: the 'date' column must be formatted as such.

If the date data is a pandas object dtype, the drop_duplicates will not work - do a pd.to_datetime first.

You've got inplace=False so you're not modifying df. You want either

 df.drop_duplicates(subset=None, keep="first", inplace=True)

or

 df = df.drop_duplicates(subset=None, keep="first", inplace=False)