isn't an advantage of APL and J that the person writing a
program/app/whatever, doesn't have to deal with the distinctions
between integer and damn near integer within the limitations of the
computer binary resolution?. In most cases this isa good thing because
close enough -given the +/- of
I suspect J interpreter didn't has the knowledge that the original string had
been .3
with .3 because what J saw was the floating point result of parsing by c
library. Ieee floating point has 15 to 16 significant digits so that 1e16 and
1e16-1 is the same number.
Perhaps one c
Inline comments follow...
> Just wondering why
>
> isgerund=: 3 : 0 :: 0
> y at .]
> 1
> )
>
> isn't an acceptable test for “gerundality”?
( I do not think so; I answered this question earlier in another post. )
>
> I also kind of agree with Bill, in the sense that J doesn’t seem to h
There's also this - http://code.jsoftware.com/wiki/Scripts/nlls - an
implementation of the Levenberg-Marquardt algo that I cribbed from someone
who copied it from APL.
On Thu, Aug 10, 2017 at 6:47 PM, Don Kelly wrote:
> I also suggest that you look at references to the use of NR for power
> syst
I also suggest that you look at references to the use of NR for power
system load flow problems which are non-linear and generally expressed
in terms of complex numbers in the polar format. These do converge
well. I haven't written one in J but one written in APL has 11 readable
lines-most o
Solving many algebraic equations in many unknowns can be done by eliminating
the unknowns one by one obtaining many algebraic equations in one unknown each,
and then solving these equations numerically.
Example: Two equations in two unknowns. 0 = (x^2)+(y^2)-16 = (y-2) .
0 = y-2 = y(y-2) = (y^2)
I find it interesting that N-R works for vectors and complex functions (and
mixes of both). Just replace all those scalar functions by their vector
equivalents:
vn=: 1 : '- n * u %. u D.1'
I added a scaling factor; it makes the convergence slower, but it fixes
problems due to precision-loss.
I agree with your contention on the basis that _3.15 = -(3.15) or _3 +_0.15
and your use of what I named fp1=:* * 1 | | does this
For the real part rp1=: ] - fp1
v
_2.375 _5.84615 _11.4 13.0028 13.0014 13 12.9986
fp1 v
_0.375 _0.84615 _0.4 0.0028 0.0014 0 0.9986
rp1 v
_2 _5 _11 13
> Is it also possible to solve a system of equation like the following
one… ?
Basically, yes.
Because not only can x = (x1,x2) be a vector, but so can y = (y1,y2) in
this adaptation of your equations:
y1 = a*(1-x1)
y2 = b*(x2-x1^2)
TABULA is an app (distributed as a JAL "addon") which employs
timing differences are sort of normal because computers have lots of
other things going on:
timespacex '+:^:(1e5 1) 2'
0.005421 12160
timespacex '+:^:(1e5 1e4) 2'
0.003509 12160
timespacex '+:^:(1e5 1e4) 2'
0.004045 12160
FYI,
--
Raul
On Thu, Aug 10, 2017 at 3:28 PM, 'Pascal Jasmin'
performance measurements imply that it doesn't duplicate it,
a =. |. 4 }. i.63
20 timespacex 'a { +: ^: (<63 ) 2'
2.168e_5 40832
20 timespacex ' +: ^: (a ) 2'
1.944e_5 38144
20 timespacex ' +: ^: (63 ) 2'
5.648e_6 1536
these results don't differ by much, but surprised that they do differ.
> yes its efficient, simply collecting the results of each iteration
> while the function is applied to the "last value".
Do you mean that
F^:(5 3) y
does not do any calculation for the '3' ?
Does that mean that the evaluation can't be parallelized, and that it
proceeds in order?
Or is it even cle
Hi there,
J looks very interesting. I have no previous experience with array
languages and, being curious, started to experiment. Now, I would
like to solve a system of non-linear equations. I could only examples
solving single equations like this one:
N=: 1 : '- u % u d. 1' NB. Adverb imple
Thanks, Joey.
Will think about whether I wish to lose the odd script before proceeding.
Mike
Please reply to mike_liz@tiscali.co.uk.
Sent from my iPad
> On 10 Aug 2017, at 16:31, Joey K Tuttle wrote:
>
> The on device update path for iOS is via the "App Store" on the menu bar
> ther
The on device update path for iOS is via the "App Store" on the menu bar there,
the rightmost item is Updates. That will show you pending updates for all your
Apps. Items appearing there have a button "UPDATE" and after invoking that and
downloading/installing the update, that will change to "OP
Hi Ruda,
Take a look at the effect of boxed arguments on the Power page of the
dictionary. http://www.jsoftware.com/help/dictionary/d202n.htm
Your example would become
F^:( On Aug 10, 2017, at 7:30 AM, Rudolf Sykora wrote:
>
> Hello,
>
> if I want to apply a function F iteratively say 0, 1,
yes its efficient, simply collecting the results of each iteration while the
function is applied to the "last value".
From: Rudolf Sykora
To: programm...@jsoftware.com
Sent: Thursday, August 10, 2017 10:30 AM
Subject: [Jprogramming] apply iteratively
Hello
Hello,
if I want to apply a function F iteratively say 0, 1, 2, 3 .. N times
on y (and see
the individual results), I know I can do
F^: (i.N) y
But is this efficient or not? I.e., does the F^:3 uses the result of previously
calculated F^:2 ? Or are the calculations completely independent (as I g
I'm still an ingenu on iPad; how do I get it to update?
Searching APPS for jsoftware reveals J701, with a rectangular open button.
No evident option to update. Selecting OPEN starts up my existing version...
Thanks,
Mike
On 10/08/2017 00:55, Eric Iverson wrote:
Joey, Patrick,
It looks as
19 matches
Mail list logo
