Extracting a Dollar Value from the Beginning of a Character Field Using Regular Expressions

[jmextratall]

Hello,

I have a variable in my dataset called 'Case_Notes' that could start with a dollar value, and I want to extract that dollar value into a new variable called 'Protected Balance' if it exists. I have been directed to perl regular expressions but am getting lost and could use some help. Here is an example of what the data looks like:

Case_Notes
$15,229.00 , 9999999999 - known fraud
" 9999999999"
$1,740.00 , 9999999999- known fraud - calling on multiple accounts
$1,155.79 , 9999999999- known fraud -male fraudster calling on female account
$14,026.81, 9999999999 - known fraud - fraudster continues to call in
$12,495.00 , 9999999999- fraudster calling on multiple accounts
$9,936.00 , 9999999999- known fraud - young female calling as ch - true person lives in OH
$9,936.00 , 9999999999- known fraud - young female calling as ch - true person lives in OH
$5,150.00 , 9999999999- known fraud - male fraudster still calling in
$2,565.67 , 9999999999- Two accounts opened and different addresses. Caller stated he send docs in already.
$5,150.00 , 9999999999- known fraud - male fraudster still calling in
$5,150.00 , 9999999999- known fraud - male fraudster still calling in
9999999999- known fraud - ani calling in IVR on multiple accounts |3974950358 - known fraud - ani calling in IVR on multiple accounts
9999999999 - fraudster calling in , true person lives in SC.
9999999999 - known fraud - female calling on multiple accounts

What I need is:

Protected_Balance

$15,229.00
-
$1,740.00
$1,155.79
$14,026.81
$9,936.00
$9,936.00
$5,150.00
$2,565.67
$5,150.00 
$5,150.00

-

-

My latest attempt was:

data want;
set have;
format ProtectedBalance dollar10.2;
ProtectedBalance = prxchange('s/.*(\$\d+).*/\1/',-1,case_notes);
where Fraud_Status = 'FRAUD';
run;

But that got me nowhere. Any help would be appreciated.

[vellad]

Try this:

data want;
   format ProtectedBalance dollar10.2;
   set have; 
   if _N_=1 then do; 
       retain patt1;
       pattern1="/[\$][,.\d]{1,}/";
       patt1=prxparse(pattern1); 
   end;
   call prxsubstr(patt1, case_notes, position1, length1); 
   if position1>0 then ProtectedBalance=substr(case_notes, position1, length1);  
   if reverse(substr(ProtectedBalance,1,1))=',' then ProtectedBalance=substr(ProtectedBalance,1, length(ProtectedBalance)-1);
 run;

[PGStats · Accepted solution]

Keep it simple:

data have;
length case_notes $200;
infile datalines truncover;
input Case_Notes $200.;
datalines;
$15,229.00 , 9999999999 - known fraud
" 9999999999"
$1,740.00 , 9999999999- known fraud - calling on multiple accounts
$1,155.79 , 9999999999- known fraud -male fraudster calling on female account
$14,026.81, 9999999999 - known fraud - fraudster continues to call in
$12,495.00 , 9999999999- fraudster calling on multiple accounts
$9,936.00 , 9999999999- known fraud - young female calling as ch - true person lives in OH
$9,936.00 , 9999999999- known fraud - young female calling as ch - true person lives in OH
$5,150.00 , 9999999999- known fraud - male fraudster still calling in
$2,565.67 , 9999999999- Two accounts opened and different addresses. Caller stated he send docs in already.
$5,150.00 , 9999999999- known fraud - male fraudster still calling in
$5,150.00 , 9999999999- known fraud - male fraudster still calling in
9999999999- known fraud - ani calling in IVR on multiple accounts |3974950358 - known fraud - ani calling in IVR on multiple accounts
9999999999 - fraudster calling in , true person lives in SC.
9999999999 - known fraud - female calling on multiple accounts
;

data want;
if not prxId then prxId + prxParse("/\$[0-9,.]+\d\b/");
set have;
call prxSubstr(prxId, Case_Notes, pos, len);
if pos > 0 then Protected_Balance = input(substr(Case_Notes, pos, len), ?? comma32.);
drop prxId pos len;
run;

proc print data=want noobs; run;

[jmextratall]

Thank you very much for this! Worked perfectly. I appreciate the help.