CODEWITHSUNDEEP
pythonpandaspandas-groupbycumsum
torchern
torchern

Reputation: 48

Count number of days in a row with weather above certain temp with pandas

So I have a dataframe of daily weather data and if the temp is above 0. e.g:

|date|temp|pos_temp|

I am trying to make a column that has a cumulative sum for days in a row with temperatures above zero. I made a one-hot encoded column with a 1 or 0 for days above 0, but I'm having trouble counting the cumulative days with a '1' in the 'positive' column.

example

date      |temp|pos_temp|
2020-04-27|1   |1       |
2020-04-28|-1  |0       | 
2020-04-29|-2  |0       |
2020-04-30|4   |1       |
2020-05-01|7   |1       |
2020-05-02|10  |1       | 
2020-05-03|14  |1       |
2020-05-04|13  |1       |

I'm trying to make a column like this:

date      |temp|pos_temp|cum_above_0
2020-04-27|1   |1       |nan
2020-04-28|-1  |0       |0
2020-04-29|-2  |0       |0
2020-04-30|4   |1       |1
2020-05-01|7   |1       |2
2020-05-02|10  |1       |3
2020-05-03|14  |1       |4
2020-05-04|13  |1       |5

Upvotes: 3

Views: 766

Answers (1)

Artur Jacques Nürnberg
Artur Jacques Nürnberg

Reputation: 86

you can iterate on every line of your data frame.

temp = pd.DataFrame({'temp':[1, -1, -2, 4, 7, 10, 14, 13]})

count = 0

for index, row in temp.iterrows():
    if row['temp'] > 0:
        count += 1
    else:
        count = 0
    temp.loc[index, 'cum_above_0'] = count



   temp   cum_above_0
0   1     1.0
1   -1    0.0
2   -2    0.0
3   4     1.0
4   7     2.0
5   10    3.0
6   14    4.0
7   13    5.0

Upvotes: 3

Related Questions