The problem is very old, affecting all extended random numbers. The
digit-by-digit test for 'too large' was flawed.
Fixed for next release.
Henry Rich
On 9/23/2017 11:16 PM, Raul Miller wrote:
Ok, but note that there's a considerable difference between !100x and !144x
Thanks,
or dowhgrading?
JVERSION
Engine: j701/2011-01-10/11:25
Library: 8.02.12
Platform: Linux 64
Installer: J802 install
*./ (!144x)>(1 ?@$ !144x)
1
*./ (!144x)>(1 ?@$ !144x)
1
*./ (!144x)>(1 ?@$ !144x)
1
*./ (!144x)>(1 ?@$ !144x)
1
On Sat, Sep 23, 2017 at 11:16 PM, Rau
Ok, but note that there's a considerable difference between !100x and !144x
Thanks,
--
Raul
On Sat, Sep 23, 2017 at 11:12 PM, Jimmy Gauvin wrote:
> I get only "all ones"
>
>*./ (!100x)>(500 ?@$ !100x)
> 1
>*./ (!100x)>(500 ?@$ !100x)
> 1
>*./ (!100x)>(500 ?@$ !100x)
> 1
>*./ (
+/(!144x)<1?@$!144x
72
+/(!144x)<1?@$!144x
84
+/(!144x)<1?@$!144x
83
JVERSION
Engine: j805/j64/darwin
Beta-9: commercial/2016-07-05T17:11:06
Library: 8.05.14
Qt IDE: 1.5.3/5.6.2
Platform: Darwin 64
Installer: J805 install
InstallPath: /users/rauldmiller/applications/j64-805
I get only "all ones"
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100x)>(500 ?@$ !100x)
1
*./ (!100
I see
*./ (!10x)>(500 ?@$ !10x)
1
*./ (!100x)>(500 ?@$ !100x)
0
*./ (!100x)>(? 500 $ !100x)
0
Thanks.
Сб, 23 сен 2017, Raul Miller написал(а):
> It's an index error because the left argument is too large.
>
> The left argument would not be too large if ? had not generated a
> value that
The last statement crashed J:
$(1e6 ?@$ !144x) A. i.144
j64-802/jconsole.sh: line 6: 7160 Killed bin/jconsole
$ j64-802/jconsole.sh
$(1e6 ?@$ !144x) A. i.144
j64-802/jconsole.sh: line 6: 28836 Killed bin/jconsole
On Sat, Sep 23, 2017 at 10:59 PM, Raul
It's an index error because the left argument is too large.
The left argument would not be too large if ? had not generated a
value that was too large.
Thanks,
--
Raul
On Sat, Sep 23, 2017 at 10:57 PM, bill lam wrote:
> It seems A. not ? that raised the index error
>
>(10^249x) A. i.144
It seems A. not ? that raised the index error
(10^249x) A. i.144
25 136 10 70 86 12 82 48 114 66 1 128 134 9 22 143 96 36 14 30
76 59 141 44 72 129 45 41 75 123 39 27 24 49 6 31 106 68 50 4 57
131 80 16 34 18 3 13 51 90 108 84 37 99 32 55 23 83 69 60 11 110
121 115 33 117 65 111 71 78 116 85 7
Please try something like $(1e6 ?@$ !144x) A. i.144
Since this is a random number generating process, the error won't
always happen in the same way.
Thanks,
--
Raul
On Sat, Sep 23, 2017 at 10:41 PM, Jimmy Gauvin wrote:
> No index error in my older J version (and what is a?):
>
>$(500 ?@$
No index error in my older J version (and what is a?):
$(500 ?@$ !144x) A. i.144
500 144
10^.249{a
10 ^. 249 { a
10^.!144x
249.744
JVERSION
Engine: j701/2011-01-10/11:25
Library: 8.02.12
Platform: Linux 64
Installer: J802 install
On Sat, Sep 23, 2017 at 10:35 PM, bill lam wrote:
>
I expected no index error, and the result from ? should be less than its
right argument.
Thanks,
--
Raul
On Sat, Sep 23, 2017 at 10:35 PM, bill lam wrote:
> What is your expected result?
>
> Сб, 23 сен 2017, Raul Miller написал(а):
>>$(500 ?@$ !144x) A. i.144
>> |index error
>>
>> After in
What is your expected result?
Сб, 23 сен 2017, Raul Miller написал(а):
>$(500 ?@$ !144x) A. i.144
> |index error
>
> After inspection, I found:
>
>10^.249{a
> 249.746
>10^.!144x
> 249.744
>
> It looks like ?@$ is sometimes generating values which are too large.
>
> FYI,
>
> --
>
$(500 ?@$ !144x) A. i.144
|index error
After inspection, I found:
10^.249{a
249.746
10^.!144x
249.744
It looks like ?@$ is sometimes generating values which are too large.
FYI,
--
Raul
--
For information about J foru
14 matches
Mail list logo
