The scattr Python package

The scattr package has one function, scattr, that provides an easy way to Set Class ATTRibutes of classes derived from pre-existing Python classes.

The scattr function takes

• a class object and
• the name of a helper script and returns a subclass called SubClassAttributes that contains the methods defined in the helper script.

Essentially, this is an easy way to add user-defined functions to classes.

Class variables, method and classes can be added dynamically:

import pandas as pd
import math
# add pi to a pandas DataFrame
pd.DataFrame.pi = math.pi

The scattr package provides a cleaner way to add new attributes, because it creates a new subclass and uses method and class defined in a separate helper script.

The scattr function only adds callable objects(methods and classes), not variables.

Pandas DataFrame example

import pandas as pd
from scattr import scattr

# create a new class that inherits from pd.DataFrame
# and includes methods defined in a 'helper.py' file
ScattrFrame = scattr(cls=pd.DataFrame, src='helper')

# instantiate the new class
df = ScattrFrame(data=pd.read_csv('risk_factors_cervical_cancer.csv'))

# test methods added from helper file
df.say_hi()

# test CowClass added from helper file
df.CowClass.say_moo()

# test method from parent class
df.head(n=1)

# confirm that df is an instance of pd.DataFrame and PydyFrame
isinstance(df, (pd.DataFrame, ScattrFrame))

# confirm that ScattrFrame is a subclass of pd.DataFrame
issubclass(ScattrFrame, pd.DataFrame)