In pandas, we can achieve the same result by using the GroupBy function:
In [6]: import pandas as pd
importnumpy as np
In [7]: goal_stats_df=pd.read_csv('champ_league_stats_semifinalists.csv')
In [27]: goal_stats_df['GoalsPerGame']= goal_stats_df['Goals']/goal_stats_df['GamesPlayed']
In [27]: goal_stats_df['GoalsPerGame']= goal_stats_df['Goals']/goal_stats_df['GamesPlayed']
In [28]: goal_stats_df
Out[28]: Club Player Goals GamesPlayedGoalsPerGame
0 Atletico Madrid Diego Costa 8 9 0.888889
1 Atletico Madrid ArdaTuran 4 9 0.444444
2 Atletico Madrid RaúlGarcía 4 12 0.333333
3 Atletico Madrid AdriánLópez 2 9 0.222222
4 Atletico Madrid Diego Godín 2 10 0.200000
5 Real Madrid Cristiano Ronaldo 17 11 1.545455
6 Real Madrid Gareth Bale 6 12 0.500000
7 Real Madrid Karim Benzema 5 11 0.454545
8 Real Madrid Isco 3 12 0.250000
9 Real Madrid Ángel Di María 3 11 0.272727
10 Bayern Munich Thomas Müller 5 12 0.416667
11 Bayern Munich ArjenRobben 4 10 0.400000
12 Bayern Munich Mario Götze 3 11 0.272727
13 Bayern Munich BastianSchweinsteiger 3 8 0.375000
14 Bayern Munich MarioMandzukić 3 10 0.300000
15 Chelsea Fernando Torres 4 9 0.444444
16 Chelsea Demba Ba 3 6 0.500000
17 Chelsea Samuel Eto'o 3 9 0.333333
18 Chelsea Eden Hazard 2 9 0.222222
19 Chelsea Ramires 2 10 0.200000
In [30]: grouped = goal_stats_df.groupby('Club')
In [17]: grouped['GoalsPerGame'].aggregate(np.max)
Out[17]: Club
Atletico Madrid 0.888889
Bayern Munich 0.416667
Chelsea 0.500000
Real Madrid 1.545455
Name: GoalsPerGame, dtype: float64
In [22]: grouped['GoalsPerGame'].apply(np.max)
Out[22]: Club
Atletico Madrid 0.888889
Bayern Munich 0.416667
Chelsea 0.500000
Real Madrid 1.545455
Name: GoalsPerGame, dtype: float64