Hi Raul,
I would use Lowe's SIFT (Scale Invariant Feature Transform) for that, see
Wikipedia. I just tried it on an "arbitrary" image: selected a part of it
and scaled that down by 0.4321 using bilinear interpolation. Then used SIFT
to extract key points from the original and the scaled-down selec
The locales may be a bit confusing, and if they are slowing down the training,
then I will definitely rethink them. The main idea is that
every layer is its own object and conducts its own forward and backward passes
during training and prediction.
Every layer, including Conv2D, LSTM, SimpleLaye
I've got a situation where I've got some large images (maybe on the
order of 4000 pixels square - though many are not square) which have
had crops made from them which are quite a bit smaller. 1200 pixels
wide and 600 high is one fairly common example.
These crops were taken from the original, and
Yes, I agree... I had hoped to draw Roger out on what he had in mind
when including those
statements in _his_ sinh function which I quoted below. The cosh
function also has non-productive
if statements.
I suspect he included them to be instructive in some way; maybe the
essay had included
I completed the addition of JQT support for cobrowser.
Update the addon and
load 'gui/cobrowser'
It will detect JQT and JHS and load the correct version.
All the best,
David Mitchell
For information about J forums see http://ww
As an experiment, I compared three versions of "sinh": the one we've been
discussing, then a version with the first "if." removed and another with
both removed. For about 2000 random arguments, they all return the same
value. Not conclusive but indicative that the lines have no effect, as one
mig
Hmm.. well...
Copy (with a single preceeding 0 added to the right arg) should work
for that, with a fixed left argument.
But note that the approach I suggested could also be made to work with
a fixed sized argument.
For example:
(,a) 6 7 8 11 12 13} 20#0
(Or you could expand 20#0...).
Thank
Um... sure?
sinh=: DP&$: : (4 : 0) " 0
if. x>10^.|y do. (1r2*_1^0>y)*x exp |y end.
if. 1<|y do. -:-/x exp y,-y end.
e=. 0.1^x
n=. e (>i.1:) y (^%!@]) 1+2*i.>.0.5*y^.e
+/(x:!.0 y) (^%!@]) 1+2x*i.n
)
Your if statements (when triggered) calculate values and
Hi -
yes, I also have difficulty tracing through the very OOP-style of coding
though I have only attempted it manually so far. By my count, the example
model for the CIFAR10 problem creates 23 namespaces, each with three to
more than twenty values, most of them scalars. Also, this style of coding
Sent from my Samsung device
Original message
From: 'Mike Day' via Programming
Date: 29/04/2019 18:33 (GMT+00:00)
To: Programming forum
Subject: [Jprogramming] Fwd: Re: extended precision essay
As may be seen, Roger Hui felt unable to advise re my private enquiry
on
As may be seen, Roger Hui felt unable to advise re my private enquiry
on one of his essays
in the J Wiki. Anyone else feel like commenting on the apparently
unproductive lines of
code?
Thanks,
Mike
Forwarded Message
Subject:Re: extended precision essay
Date: Wed,
Raul,
My current need was to understand Copy because I had a unique situation
where the size of a was fixed size. Your comments would help more if a was
variable size.
On Mon, Apr 29, 2019 at 11:51 AM Raul Miller wrote:
> (Just catching up on the weekend's email now...)
>
> I am going to ignore
Devon,
I'm not sure what I would have called my difficulty with understanding
Jon's fine jlearn system, but "de-looping" may be a good choice. To me the
difficulty involves the heavy use of locales, with which I have little
experience. This has meant that trying to trace verb calls and using debug
Yes wd was used as a way to write information to the jQT window while the
training is in operation. Without it, I would have no way of knowing the status
of the model being trained. I only considered jQT usage and not JHS or console,
so this will not work for those. If you are using JHS/jConsol
(Just catching up on the weekend's email now...)
I am going to ignore the =. here (in `b =. ,a`), because that makes
most of the rest of this make sense...
It seems to me that ,(0,.~0,.0,~0,])a is going to be as concise as you
can get, unless J gets a "padding primitive".
That said, if a is larg
Yes - internally, windows supports '/' as a path separator. But it
also supports '\' and for historical reasons uses that a lot.
Thanks,
--
Raul
On Sun, Apr 28, 2019 at 2:40 AM 'Robert Raschke' via Programming
wrote:
>
> Hi Ian,
>
> it has been quite a while since I've actively used Windows. B
Hi
I tryied to acces several sensors via j dll
I posted some months ago a bare bone example
I think if you find an example how to access gps in c you could use j dll
Actually I made an artificial horizon with accellerometers .
But does not work good
If interested I can look for it and post it.
By
I haven't never tried this. However since jandroid did not
request GPS permission in its AndroidManifest so that
I think it can never access the GPS sensor.
Mon, 29 Apr 2019, Tom Arneson написал(а):
> I'd like to access the GNSS (GPS) sensor in my android devices. Has anyone
> done this, or acces
I'd like to access the GNSS (GPS) sensor in my android devices. Has anyone done
this, or accessed any other of the android sensors with J?
--
For information about J forums see http://www.jsoftware.com/forums.htm
19 matches
Mail list logo
