SAS DATES - MIN/MAX & Overlapping

[Jayaditya]

Hi ,

I have dataset as below:

data have;
attrib to_dt format=date9. ;
attrib from_dt format=date9. ;
input IDl ID2 ID3 ID4 ID5 FROM_DT: date9. TO_DT:date9. Amount;
datalines;
10 12 100 1 4 31DEC2019 31DEC2021 100.00
10 12 100 1 4 31DEC2012 31DEC2019 100.00
10 12 101 4 4 31DEC2019 31DEC2021 200.00
10 12 101 4 4 31DEC2012 31DEC2019 200.00
10 12 103 6 4 31DEC2012 31DEC2019 300.00
10 12 103 6 4 31DEC2019 31DEC2021 300.00
10 12 104 7 4 31DEC2012 31DEC2019 400.00
10 12 104 7 4 31DEC2019 31DEC2021 400.00
11 13 105 3 6 01JUL2019 31DEC4747 500.00
11 13 105 3 6 01OCT2019 31DEC4747 500.00
11 14 106 4 4 01OCT2019 31DEC4747 600.00
11 14 106 4 4 01JUL2019 31DEC4747 700.00

;

I want as below:

Dataset 1:
10 12 100 1 4 31DEC2012 31DEC2021 100.00
10 12 101 4 4 31DEC2012 31DEC2021 200.00
10 12 103 6 4 31DEC2012 31DEC2021 300.00
10 12 104 7 4 31DEC2012 31DEC2021 400.00
11 13 105 3 6 01OCT2019 31DEC4747 500.00
11 14 106 4 4 01OCT2019 31DEC4747 600.00

Dataset 2:
11 13 105 3 6 01JUL2019 31DEC4747 500.00
11 14 106 4 4 01JUL2019 31DEC4747 700.00

1. if Amt were the same and no overlapping dates i.e. if I take the first two rows then take min(to_date) and max(from_date) 

10 12 100 1 4 31DEC2019 31DEC2021 100.00
10 12 100 1 4 31DEC2012 31DEC2019 100.00

should become 

10 12 100 1 4 31DEC2012 31DEC2021 100.00

2. if amt are same or different and if the dates are overlapping for ex: below rows

11 14 106 4 4 010CT2019 31DEC4747 600.00
11 14 106 4 4 01JUL2019 31DEC4747 700.00

Then select the first row which has latest date and write the second record to new dataset.

I tried all means but could not successed so please help in this,

Thanks,

Jayaditya

[ChrisNZ]

---

@Jayaditya wrote:

Hi ,

I have dataset as below:

data have;
attrib to_dt format=date9. ;
attrib from_dt format=date9. ;
input IDl ID2 ID3 ID4 ID5 FROM_DT: date9. TO_DT:date9. Amount; T1 t2
datalines;
10 12 100 1 4 31DEC2019 31DEC2021 100.00
10 12 100 1 4 31DEC2012 31DEC2019 100.00
10 12 101 4 4 31DEC2019 31DEC2021 200.00
10 12 101 4 4 31DEC2012 31DEC2019 200.00
10 12 103 6 4 31DEC2012 31DEC2019 300.00
10 12 103 6 4 31DEC2019 31DEC2021 300.00
10 12 104 7 4 31DEC2012 31DEC2019 400.00
10 12 104 7 4 31DEC2019 31DEC2021 400.00
11 13 105 3 6 01JUL2019 31DEC4747 500.00
11 13 105 3 6 01OCT2019 31DEC4747 500.00
11 14 106 4 4 01OCT2019 31DEC4747 600.00
11 14 106 4 4 01JUL2019 31DEC4747 700.00

;

I want as below:

Dataset 1:
10 12 100 1 4 31DEC2012 31DEC2021 100.00
10 12 101 4 4 31DEC2012 31DEC2021 200.00
10 12 103 6 4 31DEC2012 31DEC2021 300.00
10 12 104 7 4 31DEC2012 31DEC2021 400.00
11 13 105 3 6 01OCT2019 31DEC4747 500.00
11 14 106 4 4 01OCT2019 31DEC4747 600.00

Dataset 2:
11 13 105 3 6 01JUL2019 31DEC4747 500.00
11 14 106 4 4 01JUL2019 31DEC4747 700.00

1. if Amt were the same and no overlapping dates i.e. if I take the first two rows then take min(to_date) and max(from_date) 

10 12 100 1 4 31DEC2019 31DEC2021 100.00
10 12 100 1 4 31DEC2012 31DEC2019 100.00

should become 

10 12 100 1 4 31DEC2012 31DEC2021 100.00

2. if amt are same or different and if the dates are overlapping for ex: below rows

11 14 106 4 4 010CT2019 31DEC4747 600.00
11 14 106 4 4 01JUL2019 31DEC4747 700.00

Then select the first row which has latest date and write the second record to new dataset.

I tried all means but could not successed so please help in this,

Thanks,

Jayaditya

---

I colour coded the data to distribute. Is that correct?

I suppose the identiifier is ID1 to ID5?

Do you only have 2 lines per identifier?