CODEWITHSUNDEEP
pythonpandas
connor449
connor449

Reputation: 1679

How to tell if a condition is met within the next n rows

I have a df that looks like this:

event    response
0           5
0           4
1           5
0           6
0           5
0           5
1           5
0           4
0           4
0           4
0           5
0           5

If df.event indicates 1, then the event of interest has occurred. If the event occurred, then I want to see what the response is. If the response is a value greater than 5, then I want to flag the response as True. Due to a potential time lag, I want to look for response values higher than 6 throughout the 3 rows preceding each event condition.

It should look like this:

event    response      responseType
0           5             False
0           4             False
1           5             False
0           6             True
0           5             False
0           5             False
1           5             False
0           4             False
0           4             False
0           4             False
0           5             False
0           5             False

To be clear about the condition:

You can see that the first time event is 1, there is a value greater than 5 in response within the 3 rows preceding the event condition, so True is labeled at the point this condition is satisfied. For the second time the event condition is met, the response value does not exceed 5 within the next 3 rows of the event condition, so the responseType stays False.

Upvotes: 0

Views: 360

Answers (2)

SeaBean
SeaBean

Reputation: 23217

You can check current row response > 5 and current row and last 2 rows count equal 1 by using .shift() as follows:

df['responseType'] = df['response'].gt(5) & (df['event'].eq(1) |
                                             df['event'].shift(1).eq(1) |
                                             df['event'].shift(2).eq(1))


print(df)

    event  response  responseType
0       0         5         False
1       0         4         False
2       1         5         False
3       0         6          True
4       0         5         False
5       0         5         False
6       1         5         False
7       0         4         False
8       0         4         False
9       0         4         False
10      0         5         False
11      0         5         False

Upvotes: 1

Quang Hoang
Quang Hoang

Reputation: 150785

I think you want to checkout rolling:

df['responseType'] = (df.event.rolling(3).max() == 1) & df.response.gt(5)

Upvotes: 0

Related Questions