CODEWITHSUNDEEP
python-3.xpandas
uday
uday

Reputation: 6713

building class that inherits pandas DataFrame

I am trying to write a class that inherits pandas' DataFrame class for some custom data that I am working on. 

class EquityDataFrame(DataFrame):
    def __init__(self, *args, **kwargs):
        DataFrame.__init__(self, *args, **kwargs)

    def myfunc1(self,...)
        ... # does something

    def myfunc2(self, ...)
        ... # does something

In PyCharm, I get the following warning for the name of the class EquityDataFrame:

Class EquityDataFrame must implement all abstract methods 

This inspection detects when there aren't all abstract properties/methods defined in the subclass

I tried googling, but I couldn't find a useful description of this warning. Can someone help me what this warning means?

Upvotes: 6

Views: 2584

Answers (3)

Gerard G
Gerard G

Reputation: 301

Create a pandas Dataframe-like class that is chainable

from pandas import DataFrame

class PrepPandas(DataFrame):
  """Class that inherits from pandas.DataFrame then customizes it with additonal methods."""
  def __init__(self, *args, **kwargs):
    super().__init__(*args, **kwargs)

  @property
  def _constructor(self):
      """
      Creates a self object that is basically a pandas.Dataframe.
      self is a dataframe-like object inherited from pandas.DataFrame
      self behaves like a dataframe + new custom attributes and methods.
      """
      return PrepPandas

  def mult2(self):
    """
    Method to demo chanining multiple function calls. Ex: pdf.mult2().mult2()
    
    Example:
      pdf = PrepPandas({'a': [1,2,3], 'b':[4,5,6]})
      print(pdf, pdf.mult2(),  pdf.mult2().mult2(), pdf.__class__, pdf.mult2().__class__)
    """
    return self * 2

Usage Example

pdf = PrepPandas({'a': [1,2,3], 'b':[4,5,6]})  # Creates a pandas df.
pdf, pdf.mult2(),  pdf.mult2().mult2(), pdf.__class__, pdf.mult2().__class__   #  Adds a cutom mult2 function to the pdf.

Output:

Out[5]: 
 (   
    a  b
 0  1  4
 1  2  5
 2  3  6,
    a   b
 0  2   8
 1  4  10
 2  6  12,
     a   b
 0   4  16
 1   8  20
 2  12  24,
 __main__.PrepPandas,
 __main__.PrepPandas
 )

Upvotes: 0

GuiGav
GuiGav

Reputation: 161

After looking for 6 years, pylint gave me the answer :
[W0223(abstract-method), CustomDataFrame] Method '_constructor_expanddim' is abstract in class 'DataFrame' but is not overridden

And indeed implementing

@property
def _constructor_expanddim(self) -> Type["CustomDataFrame"]:
    raise NotImplementedError("Not supported for CustomDataFrames!")

made the warning disappear.

Upvotes: 3

Jenobi
Jenobi

Reputation: 468

Try something like the below

import pandas as pd

class MyDataframe(pd.DataFrame):

def __init__(self, *args, **kwargs):
    super(MyDataframe, self).__init__(*args, **kwargs)

@property
def function_1(self, x):


@property
def function_2(self):

Upvotes: 0

Related Questions