subclassing dataframe with specific columns

Question

I'm trying to create my own class of Dataframe. I would like it has some specific columns when I call it. So I do this:

from pandas import DataFrame

class MyClass(DataFrame):  

    def __init__(self):    
        super(MyClass, self).__init__(columns=['Class','Conditions'])

However, when I try to append data to it,

test = MyClass()
#test = MyClass(columns=['Class','Conditions'])

test.loc[2] = ['class2',[1, 2, 3]]

I get an error:

File "C:\Miniconda3\lib\site-packages\pandas\core\indexing.py", line 189, in setitem self._setitem_with_indexer(indexer, value)

File "C:\Miniconda3\lib\site-packages\pandas\core\indexing.py", line 451, in _setitem_with_indexer self.obj._data = self.obj.append(value)._data

File "C:\Miniconda3\lib\site-packages\pandas\core\frame.py", line 6211, in append sort=sort)

File "C:\Miniconda3\lib\site-packages\pandas\core\reshape\concat.py", line 226, in concat return op.get_result()

File "C:\Miniconda3\lib\site-packages\pandas\core\reshape\concat.py", line 428, in get_result return (cons._from_axes(new_data, self.new_axes)

File "C:\Miniconda3\lib\site-packages\pandas\core\generic.py", line 356, in _from_axes return cls(data, **kwargs)

TypeError: init() takes 1 positional argument but 2 were given

I realised if I don't declare the columns in __init__ and I assign them after the class is created, I have no problem.

However, I would like to class be created with those names as columns.

Thank very much.

I can't replicate. Can you provide a minimal reproducible example with something runnable in place of Condition('var2',op.ne,3) ? — jpp
– jpp, Commented Nov 23, 2018 at 16:50
sorry @jpp my bad. You can replace condition by a list. Like [1,2,3]. I update the example. Thank you. Regards — Ger
– Ger, Commented Nov 23, 2018 at 17:50
Does it need to be a subclass of DataFrame - unless you're going to be adding/ overriding methods etc... It kind of looks like you could just do MyDataFrame = functools.partial(DataFrame, columns=['Class', 'Condtions']) to fix the column argument for the call and use that? — Jon Clements
– Jon Clements, Commented Nov 23, 2018 at 18:01
@GermánMartínez, Your code works fine for me in v0.19.2. What version are you using? — jpp
– jpp, Commented Nov 23, 2018 at 18:11
Hi @jpp the code works, but I want the columns to be predefined when the object is constructed. Regards — Ger
– Ger, Commented Nov 27, 2018 at 11:34

gavriel · Accepted Answer · 2018-11-23 18:23:44Z

If you want to use the constructor for initialization, you have to pass the arguments to parent __init__

>>> class MyClass(DataFrame):                                                                                                                                              
        def __init__(self, *args, **kwargs):                                                                                                                                   
            kwargs['columns'] = ['Class', 'Conditions']                                                                                                                        
            super(MyClass, self).__init__(*args, **kwargs)                                                                                                                     

>>> md = MyClass({'Class': [1, 2, 3]})
>>> md
   Class Conditions
0      1        NaN
1      2        NaN
2      3        NaN

Collectives™ on Stack Overflow

subclassing dataframe with specific columns

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related