How to find max in dataframe with in a certain length in python?

Question

Here's my DataFrame:

DATE_TIME
1997-02-25 12:50:00    3238.0
1997-02-25 12:55:00    3237.5
1997-02-25 13:00:00    3237.5
1997-02-25 13:05:00    3237.5
1997-02-25 13:10:00    3239.0
1997-02-25 13:15:00    3242.0

I'm trying to find a way to find the max for every 2 intervals back and still keep in the DataFrame format.

So I would like it to look like this.

DATE_TIME
1997-02-25 12:50:00    3238.0
1997-02-25 12:55:00    3238.0
1997-02-25 13:00:00    3238.0
1997-02-25 13:05:00    3237.5
1997-02-25 13:10:00    3239.0
1997-02-25 13:15:00    3242.0

I've tried a variety of different things. Any help would be greatly appreciated.

MaxU - stand with Ukraine · Accepted Answer · 2016-09-18 21:34:43Z

1

you can use .rolling().max() method:

In [141]: df
Out[141]:
            DATE_TIME     VAL
0 1997-02-25 12:50:00  3238.0
1 1997-02-25 12:55:00  3237.5
2 1997-02-25 13:00:00  3237.5
3 1997-02-25 13:05:00  3237.5
4 1997-02-25 13:10:00  3239.0
5 1997-02-25 13:15:00  3242.0

In [142]: df['roll_3_max'] = df['VAL'].rolling(3).max().bfill()

In [143]: df
Out[143]:
            DATE_TIME     VAL  roll_3_max
0 1997-02-25 12:50:00  3238.0      3238.0
1 1997-02-25 12:55:00  3237.5      3238.0
2 1997-02-25 13:00:00  3237.5      3238.0
3 1997-02-25 13:05:00  3237.5      3237.5
4 1997-02-25 13:10:00  3239.0      3239.0
5 1997-02-25 13:15:00  3242.0      3242.0

Explanation:

In [140]: df['VAL'].rolling(3).max()
Out[140]:
0       NaN
1       NaN
2    3238.0
3    3237.5
4    3239.0
5    3242.0
Name: VAL, dtype: float64

edited Sep 18, 2016 at 21:34

answered Sep 18, 2016 at 21:28

MaxU - stand with Ukraine

212k37 gold badges402 silver badges437 bronze badges

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

Collectives™ on Stack Overflow

How to find max in dataframe with in a certain length in python?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related