how to convert diagonal format to wide format?

[bhr-q]

Hello all,

many thanks to help me to convert the first design to the second design?

Id             B       w     g

 1               10     .        .

 1                .        8     .

 1               .        .        6

2              12      .           .

2               .      15        .

2               .        .         17

.

.

.

second design:

Id              B        w          g

1               10       8          6

2              12      15         17

3

.

.

.

Thanks,

[Tom · Accepted solution]

Assuming you want the last non-missing value for each variable in the group just use the UPDATE statement.  That requires two datasets, the original data and the transactions to be applied.  But you can use your existing dataset in both places.  Just start with zero observations.

data want;
  update have(obs=0) have;
  by id;
run;

[bhr-q]

Thanks so much, that's an easy way. 

[PaigeMiller]

proc summary data=have nway;
    class id;
    var b w g;
    output out=want sum=;
run;

[bhr-q]

Another easy way. Thanks so much for sharing your info.