I am trying to create a new column 'New' that:
1) if 'y' values is different from zero, give me 'y'
2) if 'y' is equal zero, give me 'yhat' value
ds y ds yhat
0 1999-03-05 45.0 1999-03-05 37.168417
1 1999-03-06 45.0 1999-03-06 37.109215
2 1999-03-07 45.0 1999-03-07 37.049726
3 1999-03-08 45.0 1999-03-08 36.987036
4 1999-03-09 45.0 1999-03-09 36.926852
5 1999-03-10 45.0 1999-03-10 36.864715
6 1999-03-11 45.0 1999-03-11 36.771622
7 1999-03-12 45.0 1999-03-12 36.712117
8 1999-03-13 45.0 1999-03-13 36.646144
9 1999-03-14 45.0 1999-03-14 36.578244
... ... ... ... ...
7356 NaT 0 2019-04-25 8.321119
7357 NaT 0 2019-04-26 8.315796
In order to do that, I am using this function:
df['New'] = np.where(df['y']!=0, df['y'], df['yhat'])
But I get an error saying:
KeyError: 'y'
Aucun commentaire:
Enregistrer un commentaire