How to calculate mean values per group and sort the output by that mean in Pandas-CodePudding

I have a dataframe like this:

color    power
red      6
red      8
green    3
red      1
yellow   10
green    5

What I want is this:

color   mean_of_power
yellow  10
red     5
green   4

I have tried df.groupby("color")["power"].mean(), but this will give me a dataframe sorted alphabetically:

color   mean_of_power
green   4
red     5
yellow  10

How can I group a dataframe by one column (color), calculate the mean of another column (power) per group and sort the output by the value of that mean?

CodePudding user response：

You can add sort_values right after that:

df.groupby("color").power.mean().sort_values(ascending=False)

CodePudding user response：

df.groupby("color")["power"].mean().sort_values('mean_of_power')