Hi , I am looking to rotate a SAS dataset (temp3) ,basically converting a tall dataset to a fat dataset. The dataset 'represents answers to each option within a question'. The original ids for the questions and its options were not labeled in a sequential/logical  manner. I have new labels for each of the 'options' within a questionid using certain criteria represented by the variable name new_optionid. Copy-pasted below is how the dataset temp3 looks after sorting and stacking datasets by sequential order of questionid and optionid. ID questionid optionid answer Name Name_of_unit sortid new_questionid new_optionid new_sortid 267899 26 837 1 Alaska xyz 1 1 Q0101C 1 378990 27 467 0 Alaska xyz 1 2 Q0201C 2 378991 27 519 0 Alaska xyz 2 2 Q0202C 3 207320 27 520 0 Alaska xyz 3 2 Q0203C 4 207330 27 521 1 Alaska xyz 4 2 Q0204C 5 207340 27 522 1 Alaska xyz 5 2 Q0205C 6 207350 27 523 1 Alaska xyz 6 2 Q0206C 7 207360 58 170 1 Alaska xyz 7 3 Q0401C 8 Questionid=26 and its optionid=837 corresponds to new_optionid Q0101C...the first 2 digits '01' in 'Q0101C' corresponds to questionid=26 and next 2 digits '01' in 'Q0101C' corresponds to optionid=837 Questionid=27 and its optionid=467 corresponds to new_optionid Q0201C...the first 2 digits '02' in 'Q0201C' corresponds to questionid=27 and next 2 digits '01' in 'Q0201C' corresponds to optionid=467..and so on Since the dataset only houses non-missing answers, the new_optionid only contains only those answered ids . I also created a numeric field called new_sortid to use arrays with numeric data only. So, new_optionid='Q0101C' corresponds to new_sortid=1 and so on. I need to have a final dataset that looks like this, where the responses in the answer column are below each new_optionid: Name Name_of_unit Q0101C Q0201C Q0202C Q0203C Q0204C Q0205C Q0206C Q0401C Alaska xyz 1 0 0 0 1 1 1 1 I used an array (below is the code I began with)...but its not giving me what I need..could I use an array without creating a new_sortid. How can I incorporate the character value of new_optionid as column names? Please advice. data temp4; array s(9) s1-s9; retain s1-s9; set temp3; by new_questionid; s(new_sortid)=answer; if last.new_questionid then output; keep new_questionid s1-s9 name name_of_unit ; run; the above code gave me (which is making it more complicated). s1 s2 s3 s4 s5 s6 s7..etc new_questionid 1 . . . . . . 1 1 0 0 0 1 1 1 2 1 0 0 0 1 1 1 3 ... View more

Hi , I have been using the the following ods statement to output a histogram using proc sgplot. This code is contained within a macro which is placed  in another large macro and is activated when certain conditions are met...i.e: %macro example; ---------------------------------- ------------------------------- ods listing gpath = "&out." ;   / /* create PNG image file */ ods graphics on / reset=index imagename = "histogram" ; proc sgplot data=out.&summary.; histogram xyz; xaxis thresholdmin=1  thresholdmax=1; run; %mend example; Within the larger code there is a statement that says: %macro largemacro; several sas codes-------------------- ------------------- ------------------- if a in (1,2,3) then %example; ods_all_close; %mend largemacro; I ran it for different examples of data and this is the first time I noticed this problem. I am able to get the dataset '&summary' and it has the variable xyz in it. I also tried removing the ods_all_close statement but it still doesn't work. I don't think the ods_all_close statement affects ods grahics ( It was placed to ensure all output destinations are closed for subsequent program runs ). There are several such ods_all_close statement in the entire code. As I mentioned it works for other examples. I would appreciate your help. Thanks ! ... View more