topic Re: Format in SAS Programming

Format

GS2 — Tue, 02 Oct 2018 19:47:26 GMT

Using SAS 9.4

proc format;

value $RACE_DESC 'WHITE' = 'Caucasian'

'BLACK' = 'African American'

'CHINESE','PAKISTANI','OTHER' = 'Other'

'UNKNOWN','NOT APPLICABLE' = 'Unknown';
RUN;

Above is format statement. I believe it is formatted correctly. Is their a way to trim the 'white' 'black' etc. to remove any weird spaces that I am not capturing in my format. (I did try copy and pasting directly from document and it did not work that way either). Thank you

Re: Format

HB — Tue, 02 Oct 2018 20:07:58 GMT

trim(race_desc) in the data step?

Re: Format

ChrisNZ — Tue, 02 Oct 2018 21:03:25 GMT

strip() to remove leading and trailing spaces

compbl() to remove extraneous inner spaces

You might want to upcase() too,

Re: Format

Tom — Tue, 02 Oct 2018 23:49:00 GMT

No. But if you switch to using an INFORMAT instead of a FORMAT then you can use REGEXP or REGEXPE option of the INVALUE statement.

https://documentation.sas.com/?docsetId=proc&docsetVersion=9.4&docsetTarget=p1pmw90bl3jzgdn1w4202kclxtho.htm&locale=en

Re: Format

GS2 — Wed, 03 Oct 2018 12:56:35 GMT

data raw.DiagSeq_1_2;
set raw.DiagSeq_One;
if RACE = (STRIP('WHITE') OR
RACE = (STRIP('BLACK') OR
RACE = (STRIP('CHINESE') OR
RACE = (STRIP('PAKISTANI') OR
RACE = (STRIP('OTHER') OR
RACE = (STRIP('UNKNOWN') OR
RACE = (STRIP('NOT APPLICABLE');
RUN;

Above is the code I tried, it is a combination of the recommendations but it is not working. Any other thoughts?

Re: Format

ChrisNZ — Wed, 03 Oct 2018 22:46:16 GMT

Wow.

You seem to have serious misunderstandings about how programming works, and to badly need training.

I doubt the punctual help we can give here is going to take you very far.

To answer the question at hand:

proc format;
  value $wood_desc 'WHITE'                       = 'Birch'
                   'BLACK'                       = 'Ebony'
                   'CHINESE','PAKISTANI','OTHER' = 'Oak'
                   'UNKNOWN','NOT APPLICABLE'    = 'Unknown';
run;
data WANT;
  A='   WHiTE';
  B=put(strip(compbl(upcase(A))), $wood_desc.);
  putlog B=;
run;

B=Birch

I had to change the data as I find yours creepy.

As a European, such data about race/religion/etc is forbidden in many countries as its use only a few decades ago brings back very dark memories.