Remove before sales

[sascode]

Hi ,
I have a dataset like this:
ItemID sold_Date Bought_Date Indicator_Sold ;
1 . 07/01/2019 0
1 . 08/02/2019 0
1 06/01/2019 06/01/2019 1
1 . 05/01/2019 0
1 . 04/01/2019 0
2 03/31/2019 03/31/2019 1
2 . 02/01/2019 0
........................................................
In this dataset I need to do two tasks:
1.i need to remove these obs for each ItemID with bought_date that is after ItemSold date and keep only these with indicatorsold =1 or indicatorsold =0 and bought date before ItemSold date.
2.After those obs above are removed I need to create an index count, for instance for itemID 1 it will have 1,2 3,
So higher the count it means more back in time.
I appreciate your help.

[PaigeMiller · Accepted solution]

I think this explanation needs to be re-written for clarity. I think your example data set needs to be augmented with examples that meet the criteria you are discussing.

i need to remove these obs for each ItemID with bought_date that is after ItemSold date and keep only these with indicatorsold =1 or indicatorsold =0 and bought date before ItemSold date.

First, you don't have any examples in the data with bought_date that is after sold_date. Next, your instructions say what to do with bought_date after sold_date, and also what to do if bought_date is before sold_date, but your data has examples where bought_date is equal to sold_date (it's not after and it's not before) and so we don't know what should be done in that case.

[sascode]

I do understand data doesn't make sense with reality, however, I tried to ask for help based on provided variables and rows,
Thank you.

[PaigeMiller]

---

@sascode wrote:
I do understand data doesn't make sense with reality, however, I tried to ask for help based on provided variables and rows,
Thank you.

---

So does this mean you are not going to re-write the explanation to make it more clear what you want?

[sascode]

I was able to solve using another method .
Thank you for you help.

[PaigeMiller]

@sascode 

People here try to help you, but we assume you will also help us. If you have found another method, show us the solution so we can all benefit.