pandas-refract 1.2.5

Unofficial convenience functions that deal with fragmenting Pandas dataframes.

pandas-refract is an MIT licensed Python package with a simple function that allows users to divide their dataframes by the ‘Truthy’ and ‘Falseyness’ of a provided array.

Because this library deals with splitting dataframes by its values, it uses imagery of light - refract and disperse, like light entering a prism. Eventually, the goal of this package is additional features to the Pandas library that allows users to .pop rows from a dataframe where a condition is met, etc. As far as I can tell this is not possible like the below example.

Ideal case would be:

target_df = df.pop(df['target_column'] == 'target_value')
non_target_df = df

What is required now is:

target_df = df[df['target_column'] == 'target_value']
non_target_df = df[df['target_column'] != 'target_value']

With pandas-refract this becomes:

target_df, non_target_df = refract(df, df['target_column'] == 'target_value']]

Obviously, this package is not providing anything not currently possible in the current Pandas library. It does, however, add a layer of convenience for more complex slicing where you need to separate, not remove, rows by conditions.

Examples

Simplest example of current Pandas requires:

df1 = df[df.column.notnull()].reset_index(drop=True)
df2 = df[df.column.isnull()].reset_index(drop=True)

or:

df1 = df[df.column == 'test_string'].reset_index(drop=True)
df2 = df[df.column != 'test_string'].reset_index(drop=True)

With pandas-refract this becomes:

df1, df2 = refract(df, df.column.notnull(), True]

and:

df1, df2 = refract(df, df.column == test_string', True]

But you don’t have to pass it explicit boolean arrays:

data = {'a': ['', 'truthy', '', 'truthy'],
        'b': [0, 1, 2, 3]
        }

df = pd.DataFrame(data)

truthy_df, falsey_df = refract(df, df.a)

More complex examples: (where ‘a’ is Falsey and ‘b’ is an odd number)

df1, df2 = refract(df, ((~df.a) & (df.b % 2 == 1)))