topic Re: Pattern Matching with multiple values in SAS Programming

Pattern Matching with multiple values

Amit_91 — Tue, 19 May 2020 16:07:39 GMT

Hi,

I would need your help on the issue:

I have a dataset with 1000 drug names and I want to pattern match those drug names with the main Pharmacy Data set so that I can only get data for the drugs I have. Hard coding 1000 values is not an option here, so i'm looking for a dynamic solution.

Both the tables have Generic Name column and I want to pattern match on these two. I want to use wildcard operator.

Could someone please help?

Thanks!

Re: Pattern Matching with multiple values

ed_sas_member — Tue, 19 May 2020 16:17:01 GMT

Hi @Amit_91

Could you please supply some datalines in a usable form (data step) to play with and provide you with a useful solution?

Best,

Re: Pattern Matching with multiple values

ballardw — Tue, 19 May 2020 16:43:07 GMT

Please be very specific on your matching requirements.

Drugs are one area where wildcards may be a poor choice because of compound names where some names are part of several others and to some extent the composition (liquid for injection or pills for example).

Re: Pattern Matching with multiple values

Amit_91 — Tue, 19 May 2020 16:58:47 GMT

data drugs;
infile datalines dlm='+';
length generic_name $ 50;
input generic_name $ ;
datalines;
CERTOLIZUMAB
GOLIMUMAB
ABATACEPT
run;

Above are only 3 drugs but i might have 1000 drugs. And i want to match these drugs with the main pharmacy table using wildcard.

Re: Pattern Matching with multiple values

Amit_91 — Tue, 19 May 2020 17:00:11 GMT

Hey,

I want to check only oral tablets. And I want to match the drugs in my table from the main Pharmacy data set using wildcard.

Thanks,

Amit

Re: Pattern Matching with multiple values

biopharma — Tue, 19 May 2020 19:04:03 GMT

This is good but only a beginning. Can you give us similarly what is in the main pharmacy table and what you look to match? People work across a number of industries and every industry data is different. If we can't see your problem, how can we write any code to solve it? You will need to provide:

1. Code to generate Drugs dataset (you have provided)
2. Code to generate Pharmacy dataset
3. Resulting dataset

Re: Pattern Matching with multiple values

DavePrinsloo — Tue, 19 May 2020 19:11:47 GMT

Could you please give an example where you are using a wildcard. The names you give as examples do not contain wildcards!

Maybe you should consider using the COMPGED function.