Pandas: custom class as column header with multi indexing

Question

I'm trying to use objects as column headers in a multi indexed dataframe but I can't seem to get it to work. __eq__, __hash__ and __str__ only work for simple data frames.

Here is a small example:

class Signal:

    def __init__(self, name):
        self.name = name


    def __eq__(self, other):
        try:
            return self.name == other or self.name == other.name
        except AttributeError as err:
            return False

    def __str__(self):
        return str(self.name)

    def __hash__(self):
        return hash(self.name)

if __name__ == '__main__':
    import pandas as pd
    import numpy as np
    a = Signal('name')
    b = Signal('name2')
    c = Signal('something')

    data = {
        ('A', a): np.arange(2),
        ('A', b): np.ones(2),
        ('B', c): np.zeros(2)
    }

    df = pd.DataFrame(data)

    print(df)
    print('-----------')
    print(df['A'])

I also tried implementing __le__, __ge__ and __ne__. That did not do anything though. I don't really have a clue what else I could do. Anybody got some ideas?

your columns names are tuples so df['name'] will not work but if you do df[('A', 'name')] this will return you some data.. but not sure what you're trying to do here? — gyx-hh
– gyx-hh, Commented Mar 29, 2018 at 20:12
It's not. If you run the example you will see that pandas does not create the MultiIndex. df['name'] was just a type, but I should be able to just use level 0 column names like df['A']. I'm trying to create a MultiIndex but use object as column headers. — Nima Mousavi
– Nima Mousavi, Commented Mar 29, 2018 at 21:47

piRSquared · Accepted Answer · 2018-03-29 22:32:48Z

2

After defining __lt__ and __gt__

class Signal:

    def __init__(self, name):
        self.name = name


    def __eq__(self, other):
        try:
            return self.name == other or self.name == other.name
        except AttributeError as err:
            return False

    def __lt__(self, other):
        return self.name < other.name

    def __gt__(self, other):
        return self.name > other.name

    def __str__(self):
        return str(self.name)

    def __hash__(self):
        return hash(self.name)

import pandas as pd
import numpy as np
a = Signal('name')
b = Signal('name2')
c = Signal('something')

data = {
    ('A', a): np.arange(2),
    ('A', b): np.ones(2),
    ('B', c): np.zeros(2)
}

df = pd.DataFrame(data)

print(df, df['A'], sep='\n\n')

     A               B
  name name2 something
0    0   1.0       0.0
1    1   1.0       0.0

   name  name2
0     0    1.0
1     1    1.0

answered Mar 29, 2018 at 22:32

piRSquared

296k68 gold badges509 silver badges654 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Nima Mousavi Over a year ago

So it was __lt__ and __gt__ after all. I swear to god I tried this as well. Must have had a type somewhere. Thanks for the clarification.

Collectives™ on Stack Overflow

Pandas: custom class as column header with multi indexing

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related