For one of the research studies I work on, we are meant to upload data to the National Data Archive (NDA). To this end, in SAS 9.4 we are trying to create several CSV files - one for each of the survey instruments used in our interviews. NDA uploads require a row at the top of the file with information for their system to read in (in this case, A1="ndar_subject" and B1=1), the second row is the variable names, and the rest is the data. One solution I hoped for would be to print data labels AND variable names, where we assign variable1="ndar_subject" and variable2="1" for labels and leave the rest blank. We tried using a title statement in conjunction with ODS CSVALL, but that only provided us with the cell A1 requirement. Another thought would be to somehow insert a row into the CSV with A1="ndar_subject" and B1=1 and then continue with a PROC PRINT. Any solution suggestions are welcome! ... View more

Part of a macro I wrote creates a report for 1 to many patients, depending on the data. The data is first transposed, and the list of patients is put into a macro variable to be used as variables in the Proc Report. It's been requested that each patient (in record_id_list) is on their own page in the report. Is there a way to add a page break dynamically?   /* &i is a site - this report is created for each site that has data */ proc transpose data = new_saes&i out=new_saes_trans&i label=des name=redcap_var prefix=pat; id record_id; idlabel sae_1; var sae3 sae4 sae5 sae5a sae6 sae7 sae8 sae9 sae10 sae11 sae12 sae13 sae14 sae15 sae15a sae16; run; /* Get variable names for each patient to use in proc report */ %let record_id_list =; proc sql noprint; select name into :record_id_list separated by ' ' from dictionary.columns where libname='WORK' and upcase(memname)="NEW_SAES_TRANS&i" and upcase(name) like 'PAT%'; quit; /* Create reports. */ proc report data=new_saes_trans&i style(column)=[fontsize = 9pt]; column des &record_id_list; define des / " " id style={fontsize = 9pt width=480 just=left}; run;   ... View more

In a data step (Base SAS 9.4), I am trying to parse my good data from my bad data. One condition of good data is that var4 contains a string that starts with H then is followed by 4 integers and a comma. Other text may be before or after this 6-character string. Examples: Good data: H2009, V7WK48 Bad data: HV2030, V5WK24   Code I thought would work: data test2 wonkydata; set test; if find(var4,'%H____,%')=0 then output wonkydata; /*Remove data in wrong format for extracting ID*/ run; However, both good and bad data are being output to wonkydata.   This is my first time posting a question here, so if I need to provide more info please let me know. Thanks! ... View more