Custom describe or aggregate without groupby

Question

I want to use groupby.agg where my group is the entire dataframe. Put another way, I want to use the agg functionality, without the groupby. I've looked for an example of this, but can not find it.

Here's what I've done:

import pandas as pd
import numpy as np

np.random.seed([3,1415])

df = pd.DataFrame(np.random.rand(6, 4), columns=list('ABCD'))
df

def describe(df):
    funcs = dict(Kurt=lambda x: x.kurt(),
                 Skew='skew',
                 Mean='mean',
                 Std='std')
    one_group = [True for _ in df.index]
    funcs_for_all = {k: funcs for k in df.columns}
    return df.groupby(one_group).agg(funcs_for_all).iloc[0].unstack().T

describe(df)

Question

How was I supposed to have done this?

Answer 1

A small compaction of your own proposal, which I think improves readability, by exploiting that DataFrame.groupby() accepts a lambda function:

def describe(df):
    funcs = dict(Kurt=lambda x: x.kurt(),
                 Skew='skew',
                 Mean='mean',
                 Std='std')
    funcs_for_all = {k: funcs for k in df.columns}
    return df.groupby(lambda _ : True).agg(funcs_for_all).iloc[0].unstack().T

describe(df)