Simple solution: data test ; Input Week $ J05 F05 M05 A05 ; Datalines; product1 12 0 0 2 product2 0 0 0 4 product3 0 7 0 5 product4 0 10 9 8 product5 0 50 20 0 Run; data test1 (drop = J05 F05 M05 A05 i j) ; set test ; array inweek J05 F05 M05 A05 ; array outweek wk1 - wk4; do i = 1 to dim(inweek); if inweek(i) ne 0 then do j = 1 to dim(inweek); outweek(j) = inweek(i); i + 1; sumweek = sum(wk1-wk4); if i > dim(inweek) then return; end; end; run; proc print; run; ... View more

They are temporary variables set by the (in= varname) option. They indicate which table the value comes from. http://www.cpc.unc.edu/services/computer/presentations/sasclass99/merge.html ... View more

data C ; merge A(in = aval) B(in = bval); by ID ; if aval and bval; run ; The IN opperator tells you if the value is in that table. ... View more

Just a random thought, but have you given the Sounds-Like (=*) opperator a try? Also, you can use Regular expressions to match oun similar text. ... View more

An executable example. NOTE: you need to have an index on the lookup set, and you need to handle errors where there is no match. For small cases the format that Patrick refered to is great, but when you get into very large lookups this method is better. data table2(index = (depcode)); depcode=1; department='a';output; depcode=2; department='b';output; depcode=3; department='c';output; run; data table1; IndCode=1;DepCode=1;output; IndCode=2;DepCode=3;output; IndCode=3;DepCode=2;output; IndCode=4;DepCode=5;output; run; data newdata; set table1; set table2 key = depcode; if (not _iorc_ = %sysrc(_sok)) then do; Department = ' ' ; _error_ = 0; end; run; ... View more

GOOGLE "sas set key" www2.sas.com/proceedings/sugi25/25/po/25p234.pdf ftp.sas.com/samples/A60409 www.jmu.edu/docs/sasdoc/sashtml/accdb/z0612367.htm ... View more

As Scott pointed out, read the documentation, don't expect cut and paste code from a forum. Lookups using a Key are quite powerful. The quick answer to your question is that you should index your lookup dataset on the key you are using. Flip ... View more

data newdata; set table1; set table2 key = depcode; if (not _iorc_ = %sysrc(_sok)) then do; Department = ' ' ; errtxt = " "; _error_ = 0; end; run; Sorry, no parentheses needed there (Code off the top of my head may sometimes contain syntax errors). This should give you the starting point to solve your problem. Message was edited by: Flip ... View more

Simpler would be, data jl; input id $15. ; id1 = scan(id,1, '\'); id2 = scan(id, 2, '\'); id1n = input(id1, best.); id2n = input(id2, best.); cards; 16516\646465 146416 6546546\546205 516884 1654685\261656 run; ... View more

I'm not sure why you want it in a macro but if an SQL will do. proc sql; select distinct length( ) as varlong from order by varlong descending; quit; ... View more

Search on 'delimited input" you will find things such as http://support.sas.com/techsup/technote/ts673.html ... View more

Technical expertise. Team building Project management Inforcing standards In depth knowledge of subject matter. ... View more

You may want to use the RANK function to examine the apostrophe to make sure it is exactly the same character. Otherwise split the word and check for the substring before and after the apostrophe. Message was edited by: Flip ... View more

A simple SQL query can get you that: proc sql; select cats( 'min length(' , put(min(length(address)), best. -l), ')') , cats( 'max length(' , put(max(length(address)), best.-l), ')') from xxx; quit; One could also do it in a datastep : data two; set one end = eof; retain big 0; retain small 10000; size = length(address); big = max(big, size); small = min(small, size); if eof then output; run; Take your pick. ... View more

As to the role based authentication. I have done that, but you have to build your own. ... View more