HOW TO FIND 1ST NAME 2ND NAME AND SO ON...?

[Saikiran_Mamidi]

data have;

input name$30.;
cards;
sai,priya,kunta,reddy,mskhdk
sgrav,mahesg,pitre,pedda
kiran,mami,chinna
ajay,kumar
divya
;
run;

For this data  first i need use countw(name),then we will get the values of higher word name i.e 5 from our data

then i will use array.. like below

data want;

set have;

array kiran(5) $ name1-name5;

do i =1 to 5;

kiran(i) =scan(name,i,',');

end;

run;

By doing this we will get 

name1 name2.............name5

sai        priya...............mskhdk

.

.

.

like that we will get exact data;

by using countw function we can find maxnumber of words then have to follow array

But here i dont know how many maxwords are their in variable i want above data using array only;

i what the above output in single data step..

many thanks

[PaigeMiller · Accepted solution]

---

@Saikiran_Mamidi wrote:

by using countw function we can find maxnumber of words then have to follow array

But here i dont know how many maxwords are their in variable i want above data using array only;

i what the above output in single data step..

---

I consider these restrictions to be inappropriate and unnecessary, and so I don't see why you can't do it using two data steps, or not using arrays.

The obvious method is to pass through the data once, obtain the maximum number of words in a given record. Then in the second data step you know how big to make the array. Now if the data set is so HUGE that passing through it twice is not a good idea, then you are stuck with solutions that will be inefficient anyway, such as setting the array size to some large number (such as 20) even when there may only be 6 names in a row, but you won't know that 6 is the number until you pass through the data, and you can't take the risk of having an array size that is too small.

[PaigeMiller]

I should add that maybe some of the SAS Communities Hash experts can do this in a single data step; but without using hash objects I don't think it is possible.