I am extremely allergic to explicit adverbs :) However, I think the
sentence u^:_1 b. _1 provides the name of the proverb, or its linear
representation if it is unnamed (there might be easier ways). For example,
test {{ u^:_1 b. _1 }}
test
datatype test {{ u^:_1 b. _1 }}
literal
+/@:*:
Thanks Raul, Elijah & Jose for your thoughts & discussion.
It's somewhat reassuring to know that I wasn't missing anything simple and
for my purposes the current solution suffices.
Out of interest, is it easier to obtain the name of the verb that "u" in
the adverb refers to?
On Thu, May 5, 2022 a
In principle, and in general, the problem of determining whether two
(arbitrary given) verbs would produce the same result for an (arbitrary
given) argument is non-computable; otherwise, it would imply that the
halting problem is decidable (as far as I know and I can see).
In practice, I use a con
Certainly.
For example:
sum=: +/
div=: %
len=: #
mean=: sum div len
avg=: mean f.
fdot=: {{
m=. u`'' assert.3=4!:0<'u'
if. 1=L.m do.(5!:1 m)`:6 else.m`:6 end.
}}
mean f.`'' -: avg f.`''
1
mean fdot`'' -: avg fdot`''
0
FYI,
--
Raul
On Tue, May 3, 2022 at 11:56 PM Elijah Stone wro
There is a programming language-theoretic definition of equivalence (not
restricted to functions). It is quite clever, but can be boiled down to: is
there a way, within the language, to tell the difference between two values?
Because j has no hard abstractions, the question of equality is rathe
There isn't really a better way to do that test.
Conceptually, what you want is a test that determines whether two
verbs would always produce the same results for the same arguments,
but that's a problem involving infinities. It's proof territory.
That said, typically we solve this kind of proble
I want to test if a particular verb was provided to my adverb.
I came up with the solution below. Is there a better way?
myadverb = {{
res=. u {:y
if. theverb f.`'' -: u f.`'' do.
res=. ({.y) ,: res
end.
res
}}
--
For
