How to count the number of group elements with pandas

Question

I have a dataframe and just want count the number of elements in each group. I know, I can use the groupby().count() to get all the counts of all the columns, but it is too much for me, I just want the number of elements in each group. How can I do this?

Here is the example:

mydf = pd.DataFrame({"fruit":["apple","banana","apple"],"weight":[7,8,3],"price":[4,5,6]})
mydf
>>     fruit  price  weight
>> 0   apple      4       7
>> 1  banana      5       8
>> 2   apple      6       3

If I use the groupby(“fruit”).mean(), I will get the value for each column.

mydf.groupby("fruit").mean()

>>         price  weight
>> fruit                
>> apple       2       2
>> banana      1       1

But my expect output is:

>>         number_of_fruit
>> fruit                
>> apple       2  
>> banana      1

How can I do this?

Asked By: xirururu

||

Source

Answer 1

You want size to count the number of each fruit:

In [102]:
mydf.groupby('fruit').size()

Out[102]:
fruit
apple     2
banana    1
dtype: int64

Answered By: EdChum

Answer 2

The answer by EdChum is great and I just want to give an alternative solution using value_counts function:

mydf["fruit"].value_counts()

Output:

apple     2
banana    1
Name: fruit, dtype: int64

Answered By: Sam S

How to count the number of group elements with pandas

Question:

Answers: