Nice. But I would like to ask Glenn- how does he define the best result?
F.e for data table
1 1 1 0 1 1 1
1 1 1 0 0 0 0
1 1 1 0 0 0 0
0 0 0 1 1 1 1
Longest row (0), 3*3 clique 012 or even rows 1,2 as most alike?
Glenn mention his algorithmic result somewhere, how is the "best" defined?
Leo
> It testifies to importance of good test data,
> as provided in a separate message by Glenn.
>
> Another correction--now obvious.
>
> ]r=: "."0;._2'101010101 100100100 100010001 000010000 010011100 '
> 1 0 1 0 1 0 1 0 1
> 1 0 0 1 0 0 1 0 0
> 1 0 0 0 1 0 0 0 1
> 0 0 0 0 1 0 0 0 0
> 0 1 0 0 1 1 1 0 0
>
> ]j=. i,I.s{~ i=.(i.>./)+/ s=.(= >./@,) (-.=i.#r)*.(+/ . *. |:) r
> 0 2
> ]k=. I.*./ j{r
> 0 4 8
>
>
> Note (f/ . g |:) is the same as f/@:g"1/~
>
>
> --- "Glenn M. Lewis" <[EMAIL PROTECTED]> wrote:
>
>> Sorry for replying to my own post again, but after a nice shower
>> (where I do my best thinking), I think I have almost figured out what
>> Oleg wrote...
>>
>> First, he compares each row of the table with each other row with the
>> following expression:
>> ="1/~ t
>> Next, he sums up the resulting planes into a single table:
>> +/@:="1/~ t
>> He then takes the identity matrix (using the size of the original
>> table... which makes me wonder if this solution will work for a
>> non-square table) and inverts it:
>> (-.=i.1{$t)
>> He multiplies this with the sums above to remove the main diagonal
>> (which represents each row matching itself):
>> (-.=i.1{$t)*.+/@:="1/~ t
>> This now contains the number of times that each row matches other rows
>> for each column. Then he marks the maximums in this table with ones
>> and everything else with zeros:
>> (= >./@,)(-.=i.1{$t)*.+/@:="1/~ t
>> Now, things start to get a little fuzzy for me...
>> This would sum the rows:
>> +/s
>> And this would perform the logical and of the columns:
>> >./s
>> But the expression (i.>./s) seems to generate the empty list, and i
>> turns out to be zero.
>> So I think I'm lost. Can anyone please shed some more light on this
>> one for me?
>> Thanks!
>> -- Glenn
>>
>> Glenn M. Lewis wrote:
>> > Wow!!! That is very impressive! I tried pulling apart the line (by
>> executing tiny bits at a time and displaying the line in boxed
>> format), and I still don't fully understand it yet, but I was able
>> to tweak it slightly to get it to produce a result... Here's my
>> minor tweak:
>> >
>> > t
>> > 1 0 1 1 1 1 1 0
>> > 0 1 1 1 1 0 1 1
>> > 1 1 1 1 0 1 1 0
>> > 1 1 1 1 0 1 1 1
>> > 1 1 0 0 1 1 1 1
>> > 1 0 1 1 1 1 1 1
>> > 1 1 1 1 1 1 1 0
>> > 0 1 0 1 1 0 1 1
>> > ]j=. i,I.s{~i=.(i.>./)@:(+/) s=.(= >./@,)(-.=i.1{$t)*.+/@:="1/~t
>> > 0 5 6
>> > ]k=. I.*./ j{t
>> > 0 2 3 4 5 6
>> >
>> > Thanks a bunch, Oleg! Hopefully before my next birthday I will have
>> figured out what you just wrote. :-)
>> > -- Glenn
>> >
>
>
>
>
> ____________________________________________________________________________________
> Bored stiff? Loosen up...
> Download and play hundreds of games for free on Yahoo! Games.
> http://games.yahoo.com/games/front
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
