topic Re: adding record for multiple rows in SAS Programming

adding record for multiple rows

Jinson — Fri, 12 Jan 2018 07:43:32 GMT

i have fixed length file which has 5 segments and each segment in turn has 20 columns each for 1000 records.

i want to group all segments together and do maupulation and then assign them back to the records they belong.

but there is no primary key for combining the segments back to their respective record and hence wnt to make a common id for all the segments belonging to each individual record

i.e want to create something like this so that i can take each segment convert the fixed length to columns and then assign each segment to their respective record based on the record id.

record
1	address segment
v1	v2	v3	v4	v5	v6
borrower segmrnt
b1	b2	b3	b4
demog segment
2	address segment
v1	v2	v3	v4	v5	v6
borrower segmrnt
b1	b2	b3	b4
demog segment

Re: adding record for multiple rows

Kurt_Bremser — Fri, 12 Jan 2018 08:00:54 GMT

Please post your example data in a data step, as described in https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-data-AKA-generate/ta-p/258712 and https://communities.sas.com/t5/help/faqpage/faq-category-id/posting?nobounce

Re: adding record for multiple rows

Jinson — Fri, 12 Jan 2018 11:19:45 GMT

kindly find attached the sample data.

here HD BS AS RS and CS represents the 5 segments and data corresponding to them are the different fields in fixed length.

this is an example of only 1 record

Re: adding record for multiple rows

Kurt_Bremser — Fri, 12 Jan 2018 11:28:21 GMT

So it looks that you have groups of records with different record types, recognizable by the value in the first 2-character column.

Each block starts with a 'BS' and ends with a 'CD'.

Can you supply the structure for the BS, AS, RS, CR, GS, SS and CD lines? And how you expect that to end up in a SAS dataset or in a group of datasets?

Re: adding record for multiple rows

ballardw — Fri, 12 Jan 2018 16:33:35 GMT

Assuming in your actual data file each record format HD, BS, AS etc. occur on a single line then something like this skeleton program may help.

data want;
   infile "your file" dlm='|' lrecl=32000;
   informat block $2.;
   /*informats to describe other variables to be read go here*/
   input block @;
   /* each input statement below should read the variables
      for that block
   */
   Select (block);
      when ('AS')
         input ASvar1 ASvar2 ASvar3 ;
      when ('BS')
         input bsvar1 bsvar2 bsvar3 ;
      when ('CS')
         input CSvar1 CSvar2 CSvar3 ;
      when ('HD')
         input HDvar1 HDvar2 HDvar3 ;
      when ('RS')
         input RSvar1 RSvar2 RSvar3 ;
     otherwise;
   end;
run;

You would have to supply the variable names and attributes from some documentation.