Gian Medri wrote:
>
> Hi!
> I have a function:
> bestfit=:(],0{&<.&<.0.5+(1+0.01*(?30)-15)*(((+/ % #)*1+#)-(+/)))
> that calculates the n+1 element.
> bestfit 90 90 90
> 90 90 90 81
> The tacit form doesn't work properly, because the "?" is not active every
> time I call bestfit 90 90 90.
> When I use
> (],0{&<.&<.0.5+(1+0.01*(?50)-25)*(((+/ % #)*1+#)-(+/))) 90 90 90
> then the function works properly.
>
> My question is if it is possible to have a tacit verb with the "?" active.
>
> Thanks
>
> Gian Medri
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
>
>
In J, when you write (f x) in your tacit code definition, J replaces (f x)
with its value. That's fine as long as verb f is a function,
but since verb ? is not a function, every appearance of (? x) will give
you a different value.
Simply said, (>:30) in your code would mean: put 31 here, while
(?30) means: put a random number between 0 and 29 here.
Henry's solution to this is to replace (?@30) with (?@(30"_)).
How does this work?
x"_ is a constant function that returns x for any argument(s) given.
Now let's consider the following simple equation:
f@(x"_) y === f x
Is it true? Yes, if f is a function, for any noun y and x in the domain
of f . Verb ?, however, is not a function, so the Eq. breaks down.
Let's try some examples:
verb =: >:@(30"_)
verb 0
31
verb 0
31
verb 1
31
It gives us 31 every time as expected according to the Eq.
Now let's try verb ? :
verb =: ?@(30"_)
verb 1
0
verb 1
8
verb 0
6
verb 0
16
Output looks random alright. The problem solved,
another one exposed.
--
View this message in context:
http://old.nabble.com/Tacit-vs.-Explicit-tp32595211s24193p32595276.html
Sent from the J Programming mailing list archive at Nabble.com.
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
