topic SAS Concatenate no nulls in SAS Procedures

SAS Concatenate no nulls

tmm — Fri, 24 Feb 2012 21:49:16 GMT

I want to concatenate 30 columns and there are 57k rows. I can concatenate but it also includes the rows with null values and I do not want them because then my next series of code does not work right because of the null values in the concatenation. The code I have is this:

data claims13; set claims14;

readm=cat(readm1-readm30);

readm=compress(readm);

run;

The compress does not work to get rid of the null values. All 30 columns vary on which one have nulls and which ones do not. Here is an example of that: (again I have 57krows)

memberid readm1 readm2 readm3 readm4 readm5 readm6 readm7 readm8 all the way to readm30

1111 1 1 1 1

2222 1 1 1 1 1

3333 1 1 1

4444 1 1 1 1

If I use the above code it does concatenate everything but it will do it like this

memberid readm

1111 1..1.11..

2222 .11.1..11

3333 111.....

4444 ..1.1.11

I do not think those are really periods but null values and I just want them all gone so that I only have the following

memberid readm

1111 1111

2222 11111

3333 111

4444 1111

SAS Concatenate no nulls

FriedEgg — Fri, 24 Feb 2012 22:00:35 GMT

data bar;

array readm[30];

do memberid=1 to 57000;

do i=1 to dim(readm);

readm=ifn(rand('table',.75,.25)=1,1,.);

end;

output;

end;

drop i;

run;

options missing=' ';

data foo;

length readm $30;

set bar;

readm=cats(of readm1-readm30);

run;

SAS Concatenate no nulls

tmm — Mon, 27 Feb 2012 15:58:28 GMT

Almost works but it changes my memberid to 1, 2, 3 and so on and does not keep the actual memberid. I need the memberID to remain intact because that is my primary key and has to go back to other code I run later. This code is to concatenate the memberid's readm columns. Once that is done I take the original memberID along with this concatenate and have to do a code that goes back to my severity tables. This output is:(the memberids in here are made up due to HIPAA. The memberid is actual our members social securiy number)

unique memberid readm severitycode severitylevel

1111 11 1 minor

2222 1 2 moderate

3333 11111 3 major

4444 111 4 excessive

SAS Concatenate no nulls

tmm — Mon, 27 Feb 2012 16:02:29 GMT

Actually this worked to get rid of nulls and concatenate my columns

data readm.claims13; set readm.claims12;

readm=cats(readm1-readm30);

readm=compress(readm);

run;