CODEWITHSUNDEEP
pythonpandasdataframe
chessosapiens
chessosapiens

Reputation: 3409

how to add a column based on values in two previous rows in pandas

i have the following dataframe , need to add a third boolean column with 0 or 1 value, the value of the column would be 1 if the price column for both two previous day is 0.12 otherwise it would be 0

date            price    new_column
'2017-10-11'     0.13      0
'2017-10-12'     0.12      0
'2017-10-13'     0.12      0
'2017-10-14'     0.15      1
'2017-10-15'     0.13      0
'2017-10-16'     0.12      0
'2017-10-17'     0.12      0
'2017-10-18'     0.15      1

one solution can be first adding two columns that keep yesterday price and the day before yesterday price then when these two columns are both 0.12 new_column would be 1 but i am looking for a more efficient and faster solution without need to add too much extra columns

Upvotes: 2

Views: 48

Answers (1)

jezrael
jezrael

Reputation: 862611

Compare shifted columns and cast boolean mask to int:

df['new'] = ((df['price'].shift(1) == 0.12) & (df['price'].shift(2) == 0.12)).astype(int)
print (df)
           date  price  new_column  new
0  '2017-10-11'   0.13           0    0
1  '2017-10-12'   0.12           0    0
2  '2017-10-13'   0.12           0    0
3  '2017-10-14'   0.15           1    1
4  '2017-10-15'   0.13           0    0
5  '2017-10-16'   0.12           0    0
6  '2017-10-17'   0.12           0    0
7  '2017-10-18'   0.15           1    1

Upvotes: 2

Related Questions