first second third values with missing value in a row

[chimei0403]

Hi SAS experts, 

I need your advice regarding above situation.

I would like to have variables: expect1, expect2, expect3, expect4 by using y_2016, y_2017, y_2018, y_2019 variables.

For expect1 variables, I use coalesce to get this index date. Something like below.

expect1= coalesce(y_2016, y_2017, y_2018, y_2019) ;

but I would like to keep the missing values if there is no information.

(I would like to keep missing values in the row)

In this case, each row represent one patient dates.

Goal is to have t, t+1, t+2, t+3, t+4 information.

Many thanks!

data have ;
  input id y_2016 y_2017 y_2018  y_2019    expect1 expect2 expect3 expect4;
  CARDS;

a   201604  .     .         201901          201604      .        .     201901 
b   .         201708   201811 .             201708    201811     .      .
c    .         .        .     201911        201911      .        .      .
;
run;

[Tom · Accepted solution]

Figure out where to start. Then copy.

data have ;
  input id $ have1-have4;
cards;
a   201604 .      .      201901       
b   .      201708 201811 .         
c   .      .      .      201911    
d   .      .      .      .
;

data want;
 set have ;
 array old have1-have4;
 array new want1-want4;
 do start=1 to dim(old) until (not missing(old[start])); end;
 target=1;
 do index=start to dim(old);
   new[target]=old[index];
   target+1;
 end;
 drop index target;
run;

[chimei0403]

Thank you so much Tom! You are amazing!!