Re: Prime Number Generato

 BBS: Inland Empire Archive
Date: 10-18-92 (03:51)             Number: 319
From: VICTOR YIU                   Refer#: NONE
  To: ERIC B. FORD                  Recvd: NO  
Subj: Re: Prime Number Generato      Conf: (2) Quik_Bas

 -=> Quoting Eric B. Ford to Victor Yiu <=-

 > Stats:  On a 486/34 w/256K secondary cache:

 > Prime numbers up to 10000 in .3 seconds.
:*Sorry -- but it was a typo -- it was .38.

 > Note:  It was compiled under QB 4.5 and the numbers
 > are printed out using PRINT X;...

I could not read it, so your version expanded:

-------------clip
DEFINT A-Z


Last = 10000
DIM Prime(Last + 1) AS INTEGER

H = Last \ 2

T! = TIMER

FOR I = 3 TO H STEP 2
    J = 2
    WHILE J * I <= Last
        Prime(J * I) = 1
        J = J + 1
    WEND
NEXT I

FOR J = 4 TO Last STEP 4
    Prime(J) = 1
NEXT J

PRINT "2  3";

FOR I = 3 TO Last STEP 2        ' I opt.'d here: STEP 2
    IF Prime(I) = 0 THEN PRINT I;
NEXT I

PRINT
PRINT "Computed in"; TIMER - T!; "seconds"

-------------end of clip

I am very impressed -- I have not though of an algorithm similar to
yours.  My algorithm work its way up testing if it is...  Yours makes
a list of all the NON primes...  Amazing -- why didn't I think of that?

Now, down to the benchmarks.  In the QB environment, it was about the
same.  But when I compiled it, mine kept on getting .38.  But yours
fluctuated from .29 to .34.  Incredible!  It's fast.  Mine -- well --
it must be the "GUI" interface <G>.  Incredible...

I would like others to see if they could speed it up even more in QB...

... A clean desk is a sign of a cluttered desk drawer.
--- Blue Wave/RA v2.10 [NR]
 * Origin: Hard Disc Cafe / Houston Texas / (713) 589-2690 / (1:106/30.0)