topic Re: Multiple Parts of a Variable in one line of data in SAS Programming

Multiple Parts of a Variable in one line of data

angelface123ish — Sun, 18 Oct 2020 14:31:11 GMT

How could I assign each age to each corresponding name if the data looks like this?

data siblings;
input siblingnames $ age @@;

datalines;
cindy brianna lindsey steven jayden
23 20 16 22 11

;
proc print;
run;

Re: Multiple Parts of a Variable in one line of data

ballardw — Mon, 19 Oct 2020 13:44:38 GMT

First thing is to make sure any example data step recreates your current data.

Since that data step as provided generates errors then it pretty obviously does not represent your data.

Is your question how to read the data? Not obvious from the phrasing. And if your actual data has more columns or more rows the following may not work though may be extendable

data siblings;
informat name1-name5 $10.;
input name1-name5
     /age1-age5
;

datalines;
cindy brianna lindsey steven jayden
23 20 16 22 11
;

/* if you want one row per name/age*/
data want;
   set siblings;
   array n name1-name5;
   array a age1-age5;
   length name $10.;
   do i=1 to dim(n);
     name=n[i];
     age =a[i];
     output;
   end;
   keep name age;
run;

The / in the INPUT statement say "continue reading on next line".

Doing similar operations with multiple variables is the purpose of ARRAYs , to create a shorthand way of referencing the variables with an index value so multiple related values can be addressed.

The output in the second data step tells SAS when to explicitly write to the output data step, once for each name/age pair. The keep says which variables should be in the output.

If by any chance this poorly structured data came from a spreadsheet you might consider using the spreadsheet tools to transpose the rows to columns before doing anything with SAS.

Hints with datalines: do not have empty lines or leading spaces. Such may cause unexpected results.

Re: Multiple Parts of a Variable in one line of data

angelface123ish — Sun, 18 Oct 2020 15:39:18 GMT

Thank you so much for this. I was thinking about an array but wasn't quite sure how to go about it. I truly appreciate it.

Re: Multiple Parts of a Variable in one line of data

FreelanceReinh — Sun, 18 Oct 2020 15:49:09 GMT

Hi @angelface123ish,

Here is an option without arrays:

data siblings(drop=_:);
length siblingname $15;
infile datalines col=pos;
do _v=1 to 10; /* or, after an "input @;" statement: ... to 2*countw(_infile_) */
  if mod(_v,2) then do; 
    input #1 @_n siblingname @;
    _n=pos;
  end;
  else do;
    input #2 @_a age @;
    _a=pos;
    output;
  end;
end;
datalines;
cindy brianna lindsey steven jayden
23 20 16 22 11
;

Re: Multiple Parts of a Variable in one line of data

angelface123ish — Sun, 18 Oct 2020 21:12:05 GMT

Thank you so much. This is just what I wanted.