If you get 1 from either
>./ NID #/.OID
or
>/. OID #/. NID
then you can represent the connection matrix with a connection list,
which would be about the same size as your OID or NID list.
I hope this makes sense,
--
Raul
On Mon, Sep 6, 2021 at 3:53 AM Pablo Landherr <pablo.landh...@gmail.com> wrote:
>
> Thank you for your clever solutions. Unfortunately I underestimated the
> problem of the size of the data I'm processing. As nid and oid of my test
> data each have a tally of 3.16e7, the connection matrix of my data contains
> 1e15 bits, which naturally gives me a limit error. I'll have to chew
> through it in chunks using a much less elegant approach.
>
> Thanks,
> Pablo
>
> On Fri, Sep 3, 2021 at 6:59 PM Raul Miller <rauldmil...@gmail.com> wrote:
>
> > Your approach is more compact than mine, which would be a significant
> > advantage for large collections.
> >
> > Also, I'd discard the 0 value before grouping (since 0 here is an
> > artifact of the representation and not a part of the original data
> > set).
> >
> > This gives:
> >
> > OID=: 1 9 6 2 10 7 3 11 4
> > NID=: 2 10 7 3 11 12 4 8 5
> > G=: NID OID} i.1+>./NID,OID
> >
> > (</. #\)}.{:@({&G^:a:"0)i.#G
> > +---------+------+---------+
> > |1 2 3 4 5|6 7 12|8 9 10 11|
> > +---------+------+---------+
> >
> > Or, this could be sorted.
> >
> > That said, note that this only works for tree structured graphs. Which
> > is probably always going to be the case here.
> >
> > Thanks,
> >
> > --
> > Raul
> >
> > On Fri, Sep 3, 2021 at 12:53 PM Michal Wallace <michal.wall...@gmail.com>
> > wrote:
> > >
> > > Oh, Raul's version using +. (or) on the connection matrix is way nicer
> > than
> > > my version, and gets rid of my bug with directions. Do that instead. :)
> > >
> > > On Fri, Sep 3, 2021 at 12:31 PM Raul Miller <rauldmil...@gmail.com>
> > wrote:
> > >
> > > > I should note that your example connection matrix does not seem to
> > > > match the oid, nid values you displayed.
> > > >
> > > > OID=: 1 9 6 2 10 7 3 11 4
> > > > NID=: 2 10 7 3 11 12 4 8 5
> > > >
> > > > Here's the connection matrix I see represented:
> > > > ]CM=: 1 (<:OID,.NID)} 0$~,~>./OID,NID
> > > > 0 1 0 0 0 0 0 0 0 0 0 0
> > > > 0 0 1 0 0 0 0 0 0 0 0 0
> > > > 0 0 0 1 0 0 0 0 0 0 0 0
> > > > 0 0 0 0 1 0 0 0 0 0 0 0
> > > > 0 0 0 0 0 0 0 0 0 0 0 0
> > > > 0 0 0 0 0 0 1 0 0 0 0 0
> > > > 0 0 0 0 0 0 0 0 0 0 0 1
> > > > 0 0 0 0 0 0 0 0 0 0 0 0
> > > > 0 0 0 0 0 0 0 0 0 1 0 0
> > > > 0 0 0 0 0 0 0 0 0 0 1 0
> > > > 0 0 0 0 0 0 0 1 0 0 0 0
> > > > 0 0 0 0 0 0 0 0 0 0 0 0
> > > >
> > > > Note that I am using a recent J version here. In older versions of J,
> > > > that would have to be 1 (<"1<:OID,.NID)} 0$~,~>./OID,NID
> > > >
> > > > Anyways, its transitive closure would be:
> > > > (+. +./ .*~)^:_ CM
> > > > 0 1 1 1 1 0 0 0 0 0 0 0
> > > > 0 0 1 1 1 0 0 0 0 0 0 0
> > > > 0 0 0 1 1 0 0 0 0 0 0 0
> > > > 0 0 0 0 1 0 0 0 0 0 0 0
> > > > 0 0 0 0 0 0 0 0 0 0 0 0
> > > > 0 0 0 0 0 0 1 0 0 0 0 1
> > > > 0 0 0 0 0 0 0 0 0 0 0 1
> > > > 0 0 0 0 0 0 0 0 0 0 0 0
> > > > 0 0 0 0 0 0 0 1 0 1 1 0
> > > > 0 0 0 0 0 0 0 1 0 0 1 0
> > > > 0 0 0 0 0 0 0 1 0 0 0 0
> > > > 0 0 0 0 0 0 0 0 0 0 0 0
> > > >
> > > > But that does not match your suggested grouping operation (foo), so
> > > > let's assume that connections work both ways:
> > > >
> > > > ]TC=: (+. +./ .*~)^:_ CM+.|:CM
> > > > 1 1 1 1 1 0 0 0 0 0 0 0
> > > > 1 1 1 1 1 0 0 0 0 0 0 0
> > > > 1 1 1 1 1 0 0 0 0 0 0 0
> > > > 1 1 1 1 1 0 0 0 0 0 0 0
> > > > 1 1 1 1 1 0 0 0 0 0 0 0
> > > > 0 0 0 0 0 1 1 0 0 0 0 1
> > > > 0 0 0 0 0 1 1 0 0 0 0 1
> > > > 0 0 0 0 0 0 0 1 1 1 1 0
> > > > 0 0 0 0 0 0 0 1 1 1 1 0
> > > > 0 0 0 0 0 0 0 1 1 1 1 0
> > > > 0 0 0 0 0 0 0 1 1 1 1 0
> > > > 0 0 0 0 0 1 1 0 0 0 0 1
> > > >
> > > > That gives us something close to the grouping you asked for:
> > > > TC </. 1+i.#TC
> > > > +---------+------+---------+
> > > > |1 2 3 4 5|6 7 12|8 9 10 11|
> > > > +---------+------+---------+
> > > >
> > > > The ordering here is different, but if that's important we could try to
> > > > fix it:
> > > >
> > > > (/: {:@>) TC </. 1+i.#TC
> > > > +---------+---------+------+
> > > > |1 2 3 4 5|8 9 10 11|6 7 12|
> > > > +---------+---------+------+
> > > >
> > > > I hope this helps,
> > > >
> > > >
> > > > --
> > > > Raul
> > > >
> > > > On Fri, Sep 3, 2021 at 11:21 AM Pablo Landherr <
> > pablo.landh...@gmail.com>
> > > > wrote:
> > > > >
> > > > > I want to group items that are linked to each other. I tried to use
> > some
> > > > > kind of connection matrix
> > > > >
> > > > > nid =/ oid NB. an example
> > > > >
> > > > > 0 0 0 1 0 0 0 0 0
> > > > >
> > > > > 0 0 0 0 1 0 0 0 0
> > > > >
> > > > > 0 0 0 0 0 1 0 0 0
> > > > >
> > > > > 0 0 0 0 0 0 1 0 0
> > > > >
> > > > > 0 0 0 0 0 0 0 1 0
> > > > >
> > > > > 0 0 0 0 0 0 0 0 0
> > > > >
> > > > > 0 0 0 0 0 0 0 0 1
> > > > >
> > > > > 0 0 0 0 0 0 0 0 0
> > > > >
> > > > > 0 0 0 0 0 0 0 0 0
> > > > >
> > > > >
> > > > > to group items linked to each other but I can't figure out how to
> > proceed
> > > > > from there. I'm hoping someone has a trick in their toolbox to share
> > with
> > > > > me.
> > > > >
> > > > > oid is the first number of all connections and nid is the second
> > number.
> > > > So
> > > > > in this example 6 is connected to 7, 7 is connected to 12 and 12 has
> > no
> > > > > further connection. What should foo be?
> > > > >
> > > > >
> > > > > oid,.nid
> > > > >
> > > > > 1 2
> > > > >
> > > > > 9 10
> > > > >
> > > > > 6 7
> > > > >
> > > > > 2 3
> > > > >
> > > > > 10 11
> > > > >
> > > > > 7 12
> > > > >
> > > > > 3 4
> > > > >
> > > > > 11 8
> > > > >
> > > > > 4 5
> > > > >
> > > > > oid foo nid
> > > > >
> > > > > ┌─────────┬─────────┬──────┐
> > > > >
> > > > > │1 2 3 4 5│9 10 11 8│6 7 12│
> > > > >
> > > > > └─────────┴─────────┴──────┘
> > > > >
> > > > >
> > > > > Thank you,
> > > > >
> > > > > Pablo
> > > > >
> > ----------------------------------------------------------------------
> > > > > For information about J forums see
> > http://www.jsoftware.com/forums.htm
> > > > ----------------------------------------------------------------------
> > > > For information about J forums see http://www.jsoftware.com/forums.htm
> > > >
> > > ----------------------------------------------------------------------
> > > For information about J forums see http://www.jsoftware.com/forums.htm
> > ----------------------------------------------------------------------
> > For information about J forums see http://www.jsoftware.com/forums.htm
> >
> ----------------------------------------------------------------------
> For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
