https://communities.sas.com/t5/SAS-Programming/Extract-digits/m-p/45776#M9477 If I have a drug name FEIBA VH 650, I use ("/ \d{1,5}\.?\d{0,4}/") to extract digits from it.<BR /> <BR /> How to extract 651 or 1200 from FEIBA VH 651-1200? Fri, 22 Jan 2010 19:17:17 GMT SASPhile 2010-01-22T19:17:17Z https://communities.sas.com/t5/SAS-Programming/Extract-digits/m-p/45776#M9477 If I have a drug name FEIBA VH 650, I use ("/ \d{1,5}\.?\d{0,4}/") to extract digits from it.<BR /> <BR /> How to extract 651 or 1200 from FEIBA VH 651-1200? Fri, 22 Jan 2010 19:17:17 GMT https://communities.sas.com/t5/SAS-Programming/Extract-digits/m-p/45776#M9477 SASPhile 2010-01-22T19:17:17Z https://communities.sas.com/t5/SAS-Programming/Extract-digits/m-p/45777#M9478 if the format is always ddd-dddd (d for a digit) and there is no digit in strings in front of it, then you may toggle with anydigit(drugname) and index(drugname, '-'). or you can scan the string with delimitator space and '-' and see what you get from the data. Fri, 22 Jan 2010 19:56:39 GMT https://communities.sas.com/t5/SAS-Programming/Extract-digits/m-p/45777#M9478 abdullala 2010-01-22T19:56:39Z https://communities.sas.com/t5/SAS-Programming/Extract-digits/m-p/45778#M9479 without the regular expression handling,, the extraction of number strings is fairly straight forward in SAS9. Just compress(), keeping blank, dash and digits. Then scan() using blank and dash as delimiters. <BR /> I think it should be something like<BR /> scan( compress( original, ' -)', 'kd' ). 1, '- ' )<BR /> <BR /> PeterC Sat, 23 Jan 2010 09:45:21 GMT https://communities.sas.com/t5/SAS-Programming/Extract-digits/m-p/45778#M9479 Peter_C 2010-01-23T09:45:21Z