Paul Archer wrote:
>
> 3:20pm, Ramprasad wrote:
> >
> > Rafaqat Ali Chaudhry wrote:
> > >
> > > The array has at least 20K elements. This piece of code is taking up
> > > almost 100% of the CUP cycles. Would somebody please suggest some
> > > alternative solution.
> >
> > You can improve the code largely by just returning immediately as you
> > get a match
> >
> > ( dont use grep when you dont need it )
> >
> > sub check {
> > my ($l_number) = @_
> > foreach ( @numbers) {
> > return 'TRUE' if($_ == $l_number);
> > }
> > return 'FALSE';
> > }
> >
> > So every false condition would have gone thru all elements but every
> > true condition will return immediately
>
> In addition to the suggestions already given, you could sort your array, and
> then look at the middle and cut it in half (potentially several times). I
> forget the name for this technique, though...
Binary search.
> For example, you have 1000 elements, sorted. You compare your number to the
> 500th element. If it's higher, you throw away the first 500 elements. Then
> you compare your number to the 250th element, decide which half your number
> belongs in, and throw away the other half. If you just do that twice, you've
> reduced the amount of data you have to look through by a factor of four.
If the array is already sorted then using a binary search will in most
cases be faster than using a linear search however the overhead of
sorting the array first might negate that advantage. If you want to be
sure which is faster you should bechmark both methods.
John
--
use Perl;
program
fulfillment
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
