Prev: adding +6 to age
Next: How to condense multiple observations
From: sasguy on 13 Mar 2010 00:16
proc format;
value parm 74='var1'
75='var2'
76='var3';



data test;
var1=1;
var2=2;
var3=3;
x=74;
y=put(x,parm.);
z=Y;

run;
i want z=1 that means y is var1 but how can i assign the value of y
as a variable in the same dataset to another variable
Thanks in advance
From: Tom Abernathy on 13 Mar 2010 12:06
On Mar 13, 12:16 am, sasguy <addanki...(a)gmail.com> wrote:
> proc format;
>  value parm 74='var1'
>             75='var2'
> 76='var3';
>
> data test;
> var1=1;
> var2=2;
> var3=3;
> x=74;
> y=put(x,parm.);
> z=Y;
>
> run;
> i want z=1 that means y is var1  but how can i assign the value of y
> as a variable in the same dataset to another variable
> Thanks in advance

One way is to use ARRAY statement and use the value of variable as the
index.
Example:
array v var1-var3;
var1=101;
x=1;
y=v(x);
--> y=101

With the numbers in your example you could just subtract 73 from X to
get the index.
For a more complicated example you could use your format trick to
convert the actual values to the index needed for that variable.

- Tom
From: xlr82sas on 13 Mar 2010 17:20
On Mar 13, 9:06 am, Tom Abernathy <tom.aberna...(a)gmail.com> wrote:
> On Mar 13, 12:16 am, sasguy <addanki...(a)gmail.com> wrote:
>
>
>
>
>
> > proc format;
> >  value parm 74='var1'
> >             75='var2'
> > 76='var3';
>
> > data test;
> > var1=1;
> > var2=2;
> > var3=3;
> > x=74;
> > y=put(x,parm.);
> > z=Y;
>
> > run;
> > i want z=1 that means y is var1  but how can i assign the value of y
> > as a variable in the same dataset to another variable
> > Thanks in advance
>
> One way is to use ARRAY statement and use the value of variable as the
> index.
> Example:
>   array v var1-var3;
>   var1=101;
>   x=1;
>   y=v(x);
> --> y=101
>
> With the numbers in your example you could just subtract 73 from X to
> get the index.
> For a more complicated example you could use your format trick to
> convert the actual values to the index needed for that variable.
>
> - Tom- Hide quoted text -
>
> - Show quoted text -

Hi All,

I think you should try vvaluex, just change z=y to z-=vvaluex(y);

Toby Dunn had a brilliant use of vvaluex whith ods output and proc
freq tables _all_ a while ago, his solution reduced my 30 lines of
code to about three lines.
Symrefs can be very powerfull.

28745 data test;
28746 var1=1;
28747 var2=2;
28748 var3=3;
28749 x=74;
28750 y=put(x,parm.);
28751 z=Y;
28752 q=vvaluex(y);
28753 put q=;
28754 run;

Q=1
NOTE: The data set WORK.TEST has 1 ob
NOTE: DATA statement used (Total proc
real time 0.00 second
cpu time 0.01 second
 | 
Pages: 1
Prev: adding +6 to age
Next: How to condense multiple observations