Yes, this "check every other against ten" was the intention of the original author. The idea was to generate a ten sometimes.
After looking a little more, it seems
that you are performing "Generate-And-display-Num" twice in the
perform loop, but only testing for ten once.
Could be wrong, but should your logic look more like:
IDENTIFICATION DIVISION.
PROGRAM-ID. randloop.
*> cobc -x -free -ffunctions-all randloop.cob
DATA DIVISION.
WORKING-STORAGE SECTION.
1 Num PIC Z9.
1 SEED PIC 9V999999999.
PROCEDURE DIVISION.
Main.
MOVE RANDOM(SECONDS-PAST-MIDNIGHT) TO SEED.
PERFORM FOREVER
PERFORM Generate-And-Display-Num
IF Num = 10
EXIT PERFORM
END-IF
END-PERFORM
GOBACK.
Generate-And-Display-Num.
COMPUTE Num = REM(RANDOM * 100, 20).
DISPLAY Num.
On 05/30/2013 09:38 PM, Hiroki MINEMATSU wrote:
Hello, Bruce
OpenCOBOL RANDOM returned biased value.
It is look like conacatinate '0.' and int(rand()*(2**32)) .
so, you execute this.
perl -e '$a = 0; for (1..1000) {$p = int(rand() * (2 ** 31)) ;$a +=
"0.".$p; }; print $a ;print "\n";'
it will show about 347 instead of 500.
Since it is wrong, we have to correct libcob.so.
(2013/05/30 23:30), Bruce M. Axtens wrote:
Context: MinGW
$ uname -a
MINGW32_NT-6.1 MERCURY 1.0.17(0.48/3/2) 2011-04-24 23:39 i686 Msys
$ cobc -v
cobc (OpenCOBOL) 1.1.0
Copyright (C) 2001-2009 Keisuke Nishida / Roger While
Built May 29 2013 21:53:19
Packaged Feb 06 2009 10:30:55 CET
Sadly, I don't get to do much COBOL. The best I can do is the occasional
RosettaCode task.
I was looking at <http://rosettacode.org/wiki/Loops/Break#COBOL>. I
copied and pasted into a text editor, made it free-form, and started to
fiddle.
I've got as far as this:
IDENTIFICATION DIVISION.
PROGRAM-ID. Random-Nums.
DATA DIVISION.
WORKING-STORAGE SECTION.
01 Num PIC Z9.
01 SEED PIC 9V999999999.
PROCEDURE DIVISION.
Main.
MOVE FUNCTION RANDOM(FUNCTION SECONDS-PAST-MIDNIGHT) TO SEED.
PERFORM FOREVER
PERFORM Generate-And-Display-Num
IF Num = 10
EXIT PERFORM
ELSE
PERFORM Generate-And-Display-Num
END-IF
END-PERFORM
GOBACK
.
Generate-And-Display-Num.
COMPUTE Num = FUNCTION REM(FUNCTION RANDOM * 100, 20)
DISPLAY Num
.
END-PROGRAM.
The END-PROGRAM seemed to be mandatory. I wouldn't compile otherwise.
The thing that has me confused (and thus all the fiddling with RANDOM)
is that the program runs properly occasionally, and other time just gets
locked on a particular value and doesn't change.
A working runs looks like this:
$ breakloop.exe
23
67
23
21
67
18
14
17
27
11
16
25
29
26
12
23
90
27
14
20
30
29
10
but another run starts off nicely, but then gets stuck
$ breakloop.exe
25
18
29
80
14
16
26
24
26
36
18
29
54
47
29
29
29
... ad infinitum
Why is it so?
Kind regards,
Bruce.
------------------------------------------------------------------------------
Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET
Get 100% visibility into your production application - at no cost.
Code-level diagnostics for performance bottlenecks with <2% overhead
Download for free and get started troubleshooting in minutes.
http://p.sf.net/sfu/appdyn_d2d_ap1
_______________________________________________
open-cobol-list mailing list
address@hidden
https://lists.sourceforge.net/lists/listinfo/open-cobol-list
|