True... though I think that only works if you are in the 'base' locale?
Thanks,
--
Raul
On Mon, Mar 11, 2019 at 2:55 PM Henry Rich wrote:
>
> You can also set or refer to public values via a locative, ideal__
>
> Henry Rich
>
> On 3/11/2019 2:53 PM, Raul Miller wrote:
> > Also, if you really w
You can also set or refer to public values via a locative, ideal__
Henry Rich
On 3/11/2019 2:53 PM, Raul Miller wrote:
Also, if you really wanted to do the public assignment, you could do
it like this:
3 :'ideal=:y' 1 0 0
(You would need to use analogous explicit definitions to access that
Also, if you really wanted to do the public assignment, you could do
it like this:
3 :'ideal=:y' 1 0 0
(You would need to use analogous explicit definitions to access that
value, instead of your locally defined value.)
FYI,
--
Raul
On Mon, Mar 11, 2019 at 2:24 PM Henry Rich wrote:
>
> I me
I meant it's your error, not an error in J
On 3/11/2019 2:19 PM, Brian Schott wrote:
I am having a similar curious result that may be related: notice that in
the script loaded first =. is used and then =: is used . I wonder if the
problem is exacerbated by the change from a verb to a noun, also.
No, this is unrelated and is an error. A public assignment to a name
that is already privately defined is a domain error. (for your
protection, since the public value would be hidden by the private value)
Henry Rich
On 3/11/2019 2:19 PM, Brian Schott wrote:
I am having a similar curious resu
I am having a similar curious result that may be related: notice that in
the script loaded first =. is used and then =: is used . I wonder if the
problem is exacerbated by the change from a verb to a noun, also.
loadd'~user/test.ijs'
ideal =. 3{."1 = i. 4
ideal =: 1 0 0
|domain error:
Yeah, looks like I wasn't saving any significant time with that square
root stunt on the logarithm calculation.
Thanks,
--
Raul
On Mon, Mar 11, 2019 at 12:38 PM Don Guinn wrote:
>
> 6!:2 'n=:*/p^<.y^.~p=.p:i._1 p:y=.10x'
>
> 0.690272
>
> 40{.":n
>
> 6952838362417071970003075865264183883398
6!:2 'n=:*/p^<.y^.~p=.p:i._1 p:y=.10x'
0.690272
40{.":n
6952838362417071970003075865264183883398
On Mon, Mar 11, 2019 at 10:07 AM 'Mike Day' via Programming <
programm...@jsoftware.com> wrote:
> As mentioned (indirectly) in my second attempt to comment, the _ q:
> version failed (on my Win
As mentioned (indirectly) in my second attempt to comment, the _ q: version
failed (on my Windows 10 laptop, with 16Gb memory!) for 10, while yours was
ok. As, of course, you know, your approach avoids factoring by predicting the
maximum prime powers.
Cheers,
Mike
Please reply to mike_li
And, that said, after playing with this, lcmseq 10x seems to be
significantly faster than >./&.(_&q:) 1+i. 10x (at least on this
machine).
FYI,
--
Raul
P.S. if you are reading this in a context where the rest of this
thread is not available, the definition of lcmseq was:
lcmseq=: 3 : 0
Wasn’t thinking clearly though because the _ q: representation is much
nicer...
Sorry about the noise...
—
Raul
On Monday, March 11, 2019, Raul Miller wrote:
> __ q: I meant..
>
> Thanks,
>
> —
> Raul
>
> On Monday, March 11, 2019, Raul Miller wrote:
>
>> That’s a good approach.
>>
>> The _ q
__ q: I meant..
Thanks,
—
Raul
On Monday, March 11, 2019, Raul Miller wrote:
> That’s a good approach.
>
> The _ q: representation is really nice for this task.
>
> Thanks,
>
> —
> Raul
>
> On Monday, March 11, 2019, Eugene Nonko wrote:
>
>> Here's the solution I ended up using:
>>
>> _&q:^:_
That’s a good approach.
The _ q: representation is really nice for this task.
Thanks,
—
Raul
On Monday, March 11, 2019, Eugene Nonko wrote:
> Here's the solution I ended up using:
>
> _&q:^:_1 >./ _ q: >: i. 1x
>
> Just factorize to prime exponents representation, find maximums and conver
Sorry - I pressed "send" instead of "paste..." !
Second try... taking a while as my J session got clogged up with extended
hangovers and I had to force a close.
You can save a bit of time and space by deferring the conversion to
extended type:
(_&q:^:_1 >./ _ q: >: i. 1x) -: _&q:^:_1 x:>.
You can save a bit of time and space by deferring the conversion to
extended type:
On 11/03/2019 07:51, Eugene Nonko wrote:
Here's the solution I ended up using:
_&q:^:_1 >./ _ q: >: i. 1x
Just factorize to prime exponents representation, find maximums and convert
back from prime exponen
Here's the solution I ended up using:
_&q:^:_1 >./ _ q: >: i. 1x
Just factorize to prime exponents representation, find maximums and convert
back from prime exponent representation.
On Sun, Mar 10, 2019 at 2:35 PM Raul Miller wrote:
> J's extended precision integer implementation is part o
Haskell does not have any clever way to short-circuit evaluation of LCM for
arbitrary precision Integer type.
LCM is defined as follows:
lcm _ 0 = 0
lcm 0 _ = 0
lcm x y = abs ((x `quot` (gcd x y)) * y)
And GCD is implemented straightforwardly using Euclid algorithm:
g
17 matches
Mail list logo
