0
$\begingroup$

I have longitudinal transaction data of a retail store where each row is a transaction done by an individual. I would like to perform a survival analysis to analyse how long a customer will transact before churning. Here I am defining churn as someone who has not made any transaction in last 3 months from current time. I am planning to use Cox Proportional Hazard Model, it requires tenure or time_to_event parameter. What is the correct way to represent tenure in this case of churn for below data

Below is the sample of data

Id.    Visit_date.       Amount.   Tenure     Churn     Age     Income
1.     04/03/2020        500        ?           No      40      56K
1.     05/03/2020        300        ?           No      32      60K
1.     05/23/2020        800        ?           No      28      90K
1.     07/04/2020        700        ?           No      40      56K
2.     02/03/2020        500        ?          Yes      43      50K
2.     01/12/2020        300        ?          Yes      60      90K
3.     03/23/2020        800        ?           No      18      80K
4.     07/04/2020        700        ?           No      20      40K

What is the correct way to put Time_to_event or Tenure value for CoxPH model?

Improve this question
$\endgroup$
2
  • $\begingroup$ Seeme to be recurrent events, which has its own tag recurrent-events. Consider adding that tag (then you need to drop one other, as five is max number of tags.) $\endgroup$
    – kjetil b halvorsen ♦
    Commented Aug 1, 2020 at 19:06
  • $\begingroup$ @kjetilbhalvorsen Thank you for the response. I am new to survival analysis and I didn't understand what you said, can you please elaborate on your answer. $\endgroup$
    – krishna koti
    Commented Aug 3, 2020 at 5:22

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.