CODEWITHSUNDEEP
pandas
volatile
volatile

Reputation: 949

group by time and other column in pandas

I have a large pandas dataframe containing columns timestamp, name, and value

index    timestamp                     name   value
0        1999-12-31 23:59:59.000107    A      16
1        1999-12-31 23:59:59.000385    B      12
2        1999-12-31 23:59:59.000404    C      25 
3        1999-12-31 23:59:59.000704    B      15
4        1999-12-31 23:59:59.001281    A      300
5        1999-12-31 23:59:59.002211    C      20
6        1999-12-31 23:59:59.002367    C      3

I want to group by time buckets (say 20ms or 20 minutes) and name, and calculate the average value for each group.

What is the most efficient manner to do it?

Upvotes: 9

Views: 9511

Answers (1)

Gustavo Bezerra
Gustavo Bezerra

Reputation: 11034

You can use pd.Grouper, but it requires you to have the timestamps on the index. So you could try something like:

df.set_index('timestamp').groupby([pd.Grouper(freq='20Min'), 'name']).mean()

Upvotes: 19

Related Questions