topic Re: First three characters of the variable are between A and Z in SAS Programming

First three characters of the variable are between A and Z

serrld113 — Mon, 14 Jan 2019 13:52:39 GMT

Hi all,

I need need help picking up the values where the first 3 characters of the value are between A and Z. I know to do this:

substr(myvar,1,3) in ('A',......,'Z')

but don't want to spell out every letter in the alphabet.

Is there a workaround this?

Thank you

Re: First three characters of the variable are between A and Z

PeterClemmensen — Mon, 14 Jan 2019 14:40:06 GMT

like this?

data have;
input string $;
datalines;
abc234
123abc
a2c123
fghfwd
run;

data want;
   set have;
   if prxmatch('/[a-z]{3}/', string);
run;

Edit: To ignore cases, add an 'i' like this

data have;
input string $;
datalines;
Abc234
123abc
a2c123
fghfwd
run;

data want;
   set have;
   if prxmatch('/[a-z]{3}/i', string);
run;

Re: First three characters of the variable are between A and Z

PaigeMiller — Mon, 14 Jan 2019 14:13:27 GMT

You could use the NOTALPHA function.

https://documentation.sas.com/?cdcId=pgmmvacdc&cdcVersion=9.4&docsetId=fedsqlref&docsetTarget=n0xguvefrac656n1cbnaimkb0byj.htm&locale=en

Something like this:

if notalpha(myvar)>3 or notalpha(myvar)=0 then ... ;

Re: First three characters of the variable are between A and Z

novinosrin — Mon, 14 Jan 2019 15:02:50 GMT

easy,safe and fast with anyalpha


data have;
input string $;
datalines;
abc234
123abc
a2c123
fghfwd
run;

data want;
set have;
_c=0;
do pos=1 to 3;
_c+anyalpha(string,pos);
end;
if _c=6 then put  string = 'check for 1st 3 letters in A-Z is true';/*write to the log*/
drop _c;
run;

Re: First three characters of the variable are between A and Z

Tom — Mon, 14 Jan 2019 15:08:03 GMT

Do you want to test each of the three characters?

Or just that the first letter falls between A and Z?

What about lowercase letters?

Re: First three characters of the variable are between A and Z

novinosrin — Mon, 14 Jan 2019 15:23:05 GMT

@PaigeMiller You are full of brains. Thank you for that idea

data have;
input string $;
datalines;
AbC234
123abc
a2c123
fghfwd
run;

data want;
set have;
if notalpha(substr(string,1,3))=0 then  put  string = 'check for 1st 3 letters in A-Z is true';
run;

Re: First three characters of the variable are between A and Z

Ksharp — Tue, 15 Jan 2019 09:05:51 GMT

data have;
input string $;
datalines;
abc234
123abc
a2c123
fghfwd
Ads23d
BS2Csd
run;

data want;
   set have;
   if prxmatch('/[^a-z]/i', substr(string,1,3)) then delete;
run;