Get maximum weight by unique person with pandas-CodePudding

I have this data:

data = {'name':['dave','bob','barry','bob','dave','barry'],
        'age':[35,30,41,50,44,53],
        'weight':[144,158,150,187,250,197]}

I want to grab the heaviest person by weight for each name - so I only want barry to appear once with an of 53 and his weight of 197 as this barry combination is heavier than 41 age barry.

I can get it working but not return back the column of weight - here's the full code:

import pandas as pd

data = {'name':['dave','bob','barry','bob','dave','barry'],
        'age':[35,30,41,50,44,53],
        'weight':[144,158,150,187,250,197]}


df = pd.DataFrame(data)

print(df.groupby('name')['age'].max())

So, I get back this as my output:

name barry 53 bob 50 dave 44 Name: age, dtype: int64

I have tried this this:

print(df.groupby('name')['age'].max())('weight')

but it doesn't work. I need it in the right order, name, age, weight.

Thanks in advance

CodePudding user response：

One method:

df.loc[df.groupby('name')['weight'].idxmax()]

Output:

    name  age  weight
5  barry   53     197
3    bob   50     187
4   dave   44     250

Alternative:

df.sort_values('weight', ascending=False).groupby('name').head(1)
# Or                                                   , as_index=False).first()

Output:

    name  age  weight
4   dave   44     250
5  barry   53     197
3    bob   50     187