pdrepr 0.2.0

eval-able string representation of pandas objects

pdrepr

pdrepr takes a pandas DataFrame as input, and attempts to output a valid Python expression that will create an identical DataFrame. Supports multiindices for rows and columns, at least for the relatively simple cases I have tested. DataFrames with datatypes other than strings, ints and floats should work if their _repr__() method also returns a string that can be passed to eval(), resulting in a similar object.

Motivation

I was tired of having to manually construct DataFrames to be used in testing, especially the reference object to be compared with the resulting DF. With this package, such a code snipped can be created from the resulting DF.

Installation

Using poetry:

poetry add pdrepr

Using pipenv:

pipenv install pdrepr

Using pip:

pip install pdrepr

Usage

>>> from pdrepr import pdrepr

>>> pdrepr(df)
pd.DataFrame({'character': ['The Nude Organist', 'BBC continuity announcer', 'The Colonel'], 'played by': ['Terry Jones and Terry Gilliam', 'John Cleese', 'Graham Chapman']}).set_index(['character'])

>>> pd.DataFrame({'character': ['The Nude Organist', 'BBC continuity announcer', 'The Colonel'], 'played by': ['Terry Jones and Terry Gilliam', 'John Cleese', 'Graham Chapman']}).set_index(['character'])
                                              played by
character                                              
The Nude Organist         Terry Jones and Terry Gilliam
BBC continuity announcer                    John Cleese
The Colonel                              Graham Chapman