How to show all observations that include special characters

[SeanZ]

I have one variable in a dataset that include letters, numbers and special characters. Is there way I can screen out observations with ONLY letters, ONLY numbers, and ONLY special characters.

[RaviKommuri · Accepted solution]

I am presenting some simple code here, You can modify as per your exact requirement!!!

data have;                                                   

Input name $;                                                

datalines;                                                   

RAVI                                                         

1234                                                         

HARI                                                         

5678                                                         

%$#@                                                         

;                                                            

data want;                                                   

set have;                                                   

if verify(trim(left(name)),'0123456789')=0 then x = 'N';     

else if verify(trim(left(name)),'!@#$%^&*()')=0 then x = 'S';

else x = 'C';                                                

Result :

Obs    name    x

1     RAVI    C

2     1234    N

3     HARI    C

4     5678    N

5     % #     S

[Peter_C]

See the functions

ANYALPHA

NOTALPHA

ANYDIGIT

NOTDIGIT

ANYCNTRL

NOTCNTRL

[Loko]

Hello,

One solution below:

data have;
length b $ 7 ;
infile datalines truncover;
input a $200.;

if length(compress(a,,'ad')) gt 1 then b="Special";
else if length(compress(a,,'d')) gt 1 then b="Letters";
  else b="Digits";
datalines;
sdijgh68%#$kjs
436346
gsgs
#$%&
eyeryt547&&
GtreUI
;
run;