topic Re: regular expressions in SAS Programming

regular expressions

SASPhile — Wed, 22 Apr 2009 15:11:39 GMT

I'm not that good at reg exp. I tried my luck ..was not sucessful. Need your help on this.

I have a filed called "Drug_name". It will have the list of drugs. example: Paracetmol 4.8MG VIAL

The task is to extract the digits from it. in this case its 4.8.

Possible scenarios:
the number of digits vary between 1 and 4
so we need to have an or condition for number of digits d d\d d\d\d d\d\d\d

and we should consider decimal also between two digits as in the example.

the values should go to the filed called strength,.

in this case strength is 4.8

Thanks,
SASphile

Re: regular expressions

sbb — Wed, 22 Apr 2009 15:49:20 GMT

Use ANYALPHA to test if non-numeric present, then use COMPRESS function to remove chars A-Z after UPCASE, in a DATA step to assign a SAS variable, then use INPUT (convert to SAS NUMERIC) and PUT function (to convert back to a "formatted" CHARACTER type variable, displayed as a number).

Scott Barry
SBBWorks, Inc.

Re: regular expressions

deleted_user — Wed, 22 Apr 2009 16:31:01 GMT

Hello SASPhile,

Here is how your program could look like:

data T01_input;
length drug $50;
infile cards delimiter=',';
input id drug;
cards;
1,Paracetmol 4.8MG VIA
2,Paracetmol 423.23422238MG VIA
3,Paracetmol 42342.38MG VIA
4,Paracetmol x.8MG VIA
;
run;

data T02_output;
set T01_input;
length myDigits $50;
if prxmatch('/.* \d{1,4}\.\d+MG .*/',drug)
then myDigits = prxchange('s/.* (\d{1,4}\.\d+)MG .*/$1/', -1, drug);
else myDigits = "NOT FOUND";
run;

proc print data=T02_output;
run;

You would get the following output:

1 Paracetmol 4.8MG VIA 1 4.8
2 Paracetmol 423.23422238MG VIA 2 423.23422238
3 Paracetmol 42342.38MG VIA 3 NOT FOUND
4 Paracetmol x.8MG VIA 4 NOT FOUND

If you need to convert the myDigit string to number ... use the input function.

Does it help?

Yoba

Re: regular expressions

SASPhile — Thu, 23 Apr 2009 15:09:19 GMT

Thnaks yoba!