DF.groupby(col).min works, mean gets a "not implemented" error

[deanhystad]

Instead of this

mpgy = mpg.groupby("model_year").mean()["mpg"]

You need this

mpgy = mpg.groupby("model_year")["mpg"].mean()

The top one computes the mean for all the columns and selects the "mpg" column. That doesn't work when some columns are not numeric such as "name" in your file. The second one computes the mean of only the "mpg" column. You can see this in the example below were mean() is applied to mpg and weight.

import pandas as pd

df = pd.read_csv("mpg.csv")[["model_year", "mpg", "weight"]]
print(df.groupby("model_year").mean())

Output:
                  mpg       weight
model_year
70          17.689655  3372.793103
71          21.250000  2995.428571
72          18.714286  3237.714286
73          17.100000  3419.025000
74          22.703704  2877.925926
75          20.266667  3176.800000
76          21.573529  3078.735294
77          23.375000  2997.357143
78          24.061111  2861.805556
79          25.093103  3055.344828
80          33.696552  2436.655172
81          30.334483  2522.931034
82          31.709677  2453.548387