Meanwhile
http://users.ece.utexas.edu/~adnan/pike.html
;)
Thanks,
--
Raul
On Tue, Aug 26, 2014 at 12:42 AM, 'Pascal Jasmin' via Programming <
programm...@jsoftware.com> wrote:
> to expand on what you are saying,
>
>timespacex '+/`0:@.([: +./ 0 = ]) 0,~ >: i.1e5'
> 0.00329024 3.28512e6
>
to expand on what you are saying,
timespacex '+/`0:@.([: +./ 0 = ]) 0,~ >: i.1e5'
0.00329024 3.28512e6
timespacex '+`0:@.(0 = *)/ 0,~ >: i.1e5'
0.0532716 2.10253e6
the first expression sees if any item is 0, and if not, it does +/ else it
returns 0. It is 20 times faster than 2nd exp
Can't say that I like any of their code or examples.
In J, In addition to boxscan there is also ^: and lexical closures
http://www.jsoftware.com/jwiki/PascalJasmin/nonads%20revisited
/ is very powerful though
|. >: i.10
10 9 8 7 6 5 4 3 2 1
applying */ to the above not only gives the fact
The best technique for avoiding work is to not specify it in the first
place.
@. does work, of course.
:: is another option.
^: is another option.
Often, though, it's better to instead specify an operation which does the
right thing instead of "exiting early". The reason for this has to do with
What techniques or methods can be used to exit a tacit at multiple points
depending on the results of the data flow through the tacit ?.
In an explicit this is achieved by using the control statement "return." a
multiple times.
A simple case for one exit would be to use something like:
f1
p.s. you can claim that you know J very well.
;)
Thanks,
--
Raul
On Mon, Aug 25, 2014 at 8:35 PM, Jose Mario Quintana
wrote:
> Pascal wrote:
> "
> use of terms like "recursion with tail call optimization" come from lisp
> fold implementations where in order to avoid stack growth/exhaustion (
http://weblambdazero.blogspot.com/2010/07/advanced-recursion-in-newlisp.html
looks like it is describing this:
memoize=: M.
http://www.integralist.co.uk/posts/understanding-recursion-in-functional-javascript-programming/#the-solution
looks like it is describing this:
trampoline=:3 :0
try.
Pascal wrote:
"
use of terms like "recursion with tail call optimization" come from lisp
fold implementations where in order to avoid stack growth/exhaustion (/
avoids such stack abuse btw), you need to pass any accumulator (state
between function calls) as a parameter to the function. boxscan bas
Take a look at Jose's also... even more tacit (though not quite the same thing).
The essential trick is to convert the "verb on the right" to an adverb
of the form @v which lets it work just fine in an adverb train.
The "not quite the same" aspect includes reversing the order of the list.
Thanks
though I found it tricky to write ad32 without boxes due to fills, I have
written essentially the same as your last line:
reduce =: 1 : (':'; 'u boxscan (<"_1 x), < y')
except yours is better because its tacit.
(i.5) + reduce (i.10) + reduce 5 0
60 55
but manipulating the boxes directl
On Mon, Aug 25, 2014 at 2:40 PM, 'Pascal Jasmin' via Programming
wrote:
> boxscan can use any list of arguments as the x, and appends any
> shape as the y, as long x u y results in a shape "compatible with
> y" (more accurately compatible with future applications of x u
> result .
boxscan can use
/ and boxscan are indeed fold and reduce in other languages. The lisp
implementations probably the easiest to understand.
The big difference between / and boxscan is that / applied to a list implies a
requirement that all items in the list are the same shape. boxscan can use any
list of argum
I have a nitpick with your phrasing here.
Instead of saying "... use / recursively... (tail call optimization)"
I think you should say something else.
Mind you, I'm not quite sure what you should say.
And, it's a nice clever bit of code. And here's another example of how
I might use the concept
I was curious in making a chunked version of running sum for the algorithmic
savings of doing 1 modulo expression per group, but also working with packet
groups, or "bytes so far".
Randdata =. 1e7 ?@:$ 256
boxscan
=: ((&.>)/)(>@:)
an explicit version for argument to boxscan:
runsum =: 4 : 0
14 matches
Mail list logo
