seems that it will not deal with the generated number with blanks in between:

x=compress("%texttonum(text=aabbcc)");

Gerhard

On Thu, 7 Jul 2011 11:54:44 -0500, Joe Matise <snoopy369@GMAIL.COM> wrote:

>Not that I'm aware of, though you could pretty easily write one in PROC >FCMP, or even in a macro. I wrote one that almost works, but doesn't quite >seem to - I'd be curious to find out why this is erroring (it seems to error >when the numbers are returned back to SAS, but that doesn't make sense to >me). > >-Joe > >options mprint mlogic symbolgen; >%macro texttonum(text=); >%do _i = 1 %to >%sysfunc(length(&text));%eval(%sysfunc(rank(%substr(&text,&_i,1)))-96) >%end; >%mend texttonum; >data test; >x=%texttonum(text=aabbcc); >put x=; >run; > > > >On Thu, Jul 7, 2011 at 11:36 AM, Aldi Kraja <aldi@wustl.edu> wrote: > >> Hi, >> >> I want to compare similarity among two strings, but my strings are in >> letters and I want to transform them into unique numbers first before I >> continue with similarity comparisons. >> >> Is there any function in SAS that transforms ASCII text string into a >> numeric vector, where each letter based on ASCII order transforms into a >> unique number, for example a -> 1, b ->2, c->3 and s.o. >> So if I have a string of aaaabbcccc transforms into: 1111223333 >> >> TIA, >> >> Aldi >>