Quite newbie to this !
i have a dataframe that looks like this:
currentMilestone m2 SLA_M6 latedeliverydate SLA_M3 earlypickupdate
m2 2020-02-21 2020-02-18 2020-03-14 2020-02-09 2020-02-08
m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08
m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08
m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08
m1 NaT 2020-03-24 2020-02-14 2020-03-13 2020-03-18
i have written that looks like this:
def flag(data):
while data.currentMilestone== 'm1'is True:
if data.SLA_M6 > data.latedeliverydate:
return 'R'
elif (data.SLA_M3 != data.earlypickupdate) & (data.latedeliverydate <= data.SLA_M6):
return 'A'
elif (data.SLA_M3 == data.earlypickupdate) & (data.latedeliverydate >= data.earlypickupdate):
return 'G'
else:
return None
the expected output is :
currentMilestone m2 SLA_M6 latedeliverydate SLA_M3 earlypickupdate flag
m2 2020-02-21 2020-02-18 2020-03-14 2020-02-09 2020-02-08 None
m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08 None
m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08 None
m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08 None
m1 NaT 2020-03-24 2020-02-14 2020-03-13 2020-03-18 R
When i run my function i dont have any result … I mean the flag is not working properly. All rows are set to None
What wrong here ?
Advertisement
Answer
Use numpy.select for this since apply is very inefficient:
import numpy as np
cond1 = data['currentMilestone'] == 'm1'
condlist = [
(data['SLA_M6'] > data['latedeliverydate']) & cond1,
(data['SLA_M3'] != data['earlypickupdate']) & (data['latedeliverydate'] <= data['SLA_M6']) & cond1,
(data['SLA_M3'] == data['earlypickupdate']) & (data['latedeliverydate'] >= data['earlypickupdate']) & cond1
]
choicelist = ['R', 'A', 'G']
data['flag'] = np.select(condlist, choicelist, default=None)
[out]
currentMilestone m2 SLA_M6 latedeliverydate SLA_M3 earlypickupdate flag 0 m2 2020-02-21 2020-02-18 2020-03-14 2020-02-09 2020-02-08 None 1 m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08 None 2 m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08 None 3 m2 2020-02-21 2020-02-18 2020-02-14 2020-02-09 2020-02-08 None 4 m1 NaT 2020-03-24 2020-02-14 2020-03-13 2020-03-18 R