It happened that comparison of two pandas dataframes could be quite handy for test automation. I picked up an original idea somewhere on Internet and modified it for my needs.
The DfComparator takes two pandas dataframes and returns an empty dataframe in case if those two dataframes are equal or a
dataframe that contains all found differences. Result dataframe has an index that identifies the location of found
differences and two coloums left_df and right_df with values that belong to corresponding dataframes.
I may add tolerance functionality later. As far as I could see it needs:
- examination of column value types, and
- some clear logic for comparision of ascii values
Apparently, if you work with static dtypes, adding tolerance for % or absolute value differences is not such a difficult task.
The code was tested with Python 3.6. To see how it works, please, run a unit test in tests folder.