find lowest value in multple columns

[Riteshdell]

Hi All,

I have 3 values corresponsing to each ID.

I am looking for output as lowest V column corresponding ID.

In the output of Final column, Value should be picked always on V3 , V2, V1 sequence priority.

If all three is not null then V3

if V1 and V2 is not null  and V3 is nullthen V2

if v2 and v3 is null and V1 is not null then v1

note : ID is duplicated in rows.

Source :

data test;
infile datalines dlm=' ' dsd;
input id $ V1 $ V2 $ V3 $;
datalines;
1 x y z
2 x y
3 x
1 x y
3 x y z
;
run;

expecting output :

ID Final

1    z

2    y

3    x

1    y

3    z

[PaigeMiller]

---

@Riteshdell wrote:

Hi All,

I have 3 values corresponsing to each ID.

I am looking for output as lowest V column corresponding ID.

In the output of Final column, Value should be picked always on V3 , V2, V1 sequence priority.

If all three is not null then V3

---

if not missing(v1) and not missing(v2) and not missing(v3) then output=v3;

if V1 and V2 is not null  and V3 is nullthen V2

if v2 and v3 is null and V1 is not null then v1

These are obvious modifications of the code above.

[Astounding]

An easy-to-understand approach:

final = v3;

if final = " " then final = v2;

if final = " " then final = v1;

[gamotte]

Hello,

Another solution using an array and coalescec function to retrieve the first non missing element :

data test;
	infile datalines dlm=' ' dsd;
	input id $ V1 $ V2 $ V3 $;
	length final $1.;
	array v(*) v3-v1;
	final=coalescec(of v(*));
	datalines;
1 x y z
2 x y
3 x
1 x y
3 x y z
;
run;