Re: ADDING ROWS FOR MISSING ORDER VARIABLES

1. Using PROC SQL or PROC SORT, create a list of the unique ORD1 values.

2. Create a table containing all of the possible ORD2/LABEL1 pairs, with
zeroes for TRT1.

3. Using PROC SQL, join #1 and #2.

4. Code a DATA step with an UPDATE statement. Use #3 as the master, WONG
as the transaction data set, and ORD1 and ORD2 as the BY variables.

On Wed, 12 Apr 2006 14:09:44 -0400, Rathindronath <mehedisas@YAHOO.COM>
wrote:

>I have a following dataset named WONG:
>
>ORD1 ORD2 LABEL1 TRT1
> 2 1 Safety Population 44(100.0)
> 2 2 10-17 years 1(2.3)
>
>I should have Value 1, 2, 3 & 4 for the variable ORD2. So I need to check
>first if ORD2 variable has all the four required value if it doesn’t (like
>the WONG dataset above) then there should be 2 ( depending how many values
>are missing: it can be 1 or 2 or 3 or even all of them) more row added
>with value 3 and 4 for ORD2 variable.
>After checking the values for ORD2 variable and adding the rows, WONG
>dataset should be all below:
>
>
>ORD1 ORD2 LABEL1 TRT1
> 2 1 Safety Population 44(100.0)
> 2 2 10-17 years 1( 2.3)
> 2 3 17-64 years 0( 0.0)
> 2 4 64-85 years 0( 0.0)
>
>Thanks in advance for your kind help.