Append dates horizontally

[carl_miles]

cust_id   date1                date2                 date3

1001        15JAN2023    20FEB2023     30MAR2023

I want an ouput like 

cust_id   date1                date2                 date3                  date 4

1001        15JAN2023    20FEB2023     30MAR2023    15JAN2023

1001        15JAN2023    20FEB2023     30MAR2023    20FEB2023 

if you guys see date1 and date2 are appended against date3 

like that i want an output 
i have dataset of 15000 rows
kindly guide me here 

[PeterClemmensen]

Do you want a third row with the value of date3 also?

[carl_miles]

No

[PaigeMiller]

---

@carl_miles wrote:

cust_id   date1                date2                 date3

1001        15JAN2023    20FEB2023     30MAR2023

I want an ouput like 

cust_id   date1                date2                 date3                  date 4

1001        15JAN2023    20FEB2023     30MAR2023    15JAN2023

1001        15JAN2023    20FEB2023     30MAR2023    20FEB2023 

if you guys see date1 and date2 are appended against date3 

like that i want an output 
i have dataset of 15000 rows
kindly guide me here 

---

I feel like I don't understand where date4 comes from, or why it is needed. I don't understand how many rows will be created from this initial one row data set. What happens if there are more CUST_ID values, then what?

[carl_miles]

ITS FINE

[PaigeMiller]

---

@carl_miles wrote:
ITS FINE

---

This extreme level of brevity does not help me understand the problem.

[mkeintz]

You can see the lack of specific suggestions in the responses to your posting.

I think this is primarily due to the absence of a clear description (or of any description at all) of your objective.

I suspect for every incoming observation, you want to generate two observations, one with DATE4=DATE1, and the second with DATE4=DATE2.  But that notion is nothing more than an attempt at mindreading, a skill I do not trust.

If you can clearly state what you want to dio you would get a lot more help.  Please help us help you.

[ballardw]

This accomplishes what you ask for the shown values.

First is a data step to actually have values.

data have;
 input cust_id date1 :date9. date2 :date9.  date3 :Date9.;
 format date: date9.;
datalines;
1001 15JAN2023 20FEB2023 30MAR2023
;

data want;
   set have;
   array d (*) date1 date2;
   do i= 1 to dim(d);
      date4=d[i];
      output;
   end;
   format date4 date9.;
   drop i;
run;

Why? I don't see any practical use.

What do you intend to do with the resulting set? It may be this step just adds complications on the way to your goal.