How to apply different aggregation functions to same column by using pandas Groupby

Question

It is clear when doing

 data.groupby(['A','B']).mean()

We get something multiindex by level 'A' and 'B' and one column with the mean of each group

how could I have the count(), std() simultaneously ?

so result looks like in a dataframe

A   B    mean   count   std

Please go through - stackoverflow.com/questions/15322632/… . Your question seems almost similar to this one. — Pralhad Narsinh Sonar
– Pralhad Narsinh Sonar, Commented Jun 5, 2015 at 19:58
@PralhadNarsinhSonar I'd say that question is different, the OP here is asking how to apply multiple different functions at once, not to generate multiple columns from a single function — EdChum
– EdChum, Commented Jun 5, 2015 at 20:00

EdChum · Accepted Answer · 2015-06-05 20:08:31Z

3

The following should work:

data.groupby(['A','B']).agg([pd.Series.mean, pd.Series.std, pd.Series.count])

basically call agg and passing a list of functions will generate multiple columns with those functions applied.

Example:

In [12]:

df = pd.DataFrame({'a':np.random.randn(5), 'b':[0,0,1,1,2]})
df.groupby(['b']).agg([pd.Series.mean, pd.Series.std, pd.Series.count])
Out[12]:
          a                
       mean       std count
b                          
0 -0.769198  0.158049     2
1  0.247708  0.743606     2
2 -0.312705       NaN     1

You can also pass the string of the method names, the common ones work, some of the more obscure ones don't I can't remember which but in this case they work fine, thanks to @ajcr for the suggestion:

In [16]:
df = pd.DataFrame({'a':np.random.randn(5), 'b':[0,0,1,1,2]})
df.groupby(['b']).agg(['mean', 'std', 'count'])

Out[16]:
          a                
       mean       std count
b                          
0 -1.037301  0.790498     2
1 -0.495549  0.748858     2
2 -0.644818       NaN     1

edited Jun 5, 2015 at 20:08

answered Jun 5, 2015 at 19:56

EdChum

397k204 gold badges836 silver badges583 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Alex Riley Over a year ago

To save a bit of typing it's possible to pass just strings of the function names to agg, for example agg(['mean', 'std', 'count']) (at least in newer versions of Pandas).

EdChum Over a year ago

yeah I know but sometimes it works and sometimes it doesn't and I can never remember which ones work so I like to be explicit for clarity, will update though, thanks for the suggestion

Collectives™ on Stack Overflow

How to apply different aggregation functions to same column by using pandas Groupby

1 Answer 1

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related