pandas aggregate function in groupby - default option?

Question

I have the following dataset (df). I want to groupby it using brand as my index, get the mean of workers and value columns and the first count of provider column.

brand   workers value   provider
H&M      322    56         mark
H&M      450    433        mark
Lindex  678     233        luke
Lindex  543     456        luke
Levi    234     32         chris
Levi    789     12         chris

Now I can

df = df.groupby('brand')['workers', 'value', 'provider'].agg({'workers': mean,  'value':mean, 'provider' : first).reset_index()

but consider that my real dataset as way more columns I want to take the mean and I don't want to specify each of them, is there a better way of declaring a default function?

Sort of "take the mean of all the non string columns and the first observation of the string columns?"

Answer 1

No, but it isn't that hard to write some code to do it for you.

f = dict.fromkeys(df, 'mean')
f.update(
    dict.fromkeys(df.columns[df.dtypes.eq(object)], 'first'))

print(f)
{'brand': 'first', 'provider': 'first', 'value': 'mean', 'workers': 'mean'}

You then pass f to agg.

df = df.groupby('brand')['workers', 'value', 'provider'].agg(f)

If you want to reset the index, you will have to remove the grouper from f.

del f['brand']
df = df.groupby('brand', as_index=False)['workers', 'value', 'provider'].agg(f)