I think Barry mean that you can represent the (x,y) pair as a single number like (max(X)-min(X))*(Y-min(Y))+X-min(X) or so, but I don't see how it would be helpful.
2018-05-02 0:20 GMT+02:00, Roman Fleysher: > Dear Barry, > > The statement about the square is not obvious to me. The requirements on > counts in x and y are different. > > I also imagine answer could be two or several non-overlapping "rectangles". > "Rectangles" will not be densely filled with dots, they might have empty > spots either because the points were never on the list or were eliminated. > > Roman > > ________________________________________ > From: sqlite-users [sqlite-users-boun...@mailinglists.sqlite.org] on behalf > of Barry Smith [smith.bar...@gmail.com] > Sent: Tuesday, May 01, 2018 6:12 PM > To: SQLite mailing list > Subject: Re: [sqlite] probably recursive? > > Well those constraints simplify your problem. > > In the resultant dataset, the largest X and Y values will be equal, and the > largest X will have and entry for every coordinate from (X, 1) to (X, X). > Likewise the largest Y will have an entry for every coordinate from (1, Y) > to (Y, Y). Basically you'll have two lines from the axes, drawing a square. > All points outside that square will be culled, all points on and inside the > square will be kept. > > Since you know that, you now have a one dimensional problem to solve. It > still seems a little recursive to me, but it should be easier because you > only need to find a single number (which you can then plug into a delete > statement). > > If my statement about the square is not obvious to prove in your head I can > try write a proof for that but I'm not much good at proofs. > >> On 2 May 2018, at 7:27 am, Roman Fleysher >> wrote: >> >> Pairs (x,y) do not repeat. >> >> Actual x and y are positive integers, but I do not see how being positive >> can be relevant. Integer is important for sorting/comparison. >> >> >> Roman >> >> ________________________________________ >> From: sqlite-users [sqlite-users-boun...@mailinglists.sqlite.org] on >> behalf of Barry Smith [smith.bar...@gmail.com] >> Sent: Tuesday, May 01, 2018 5:23 PM >> To: SQLite mailing list >> Subject: Re: [sqlite] probably recursive? >> >> Is there a uniqueness constraint on your initial data? Can the same >> coordinate be listed multiple times? >> >> Is there a requirement that X > 0 and Y > 0? >> >>>> On 2 May 2018, at 3:35 am, Simon Slavin wrote: >>>> >>>> On 1 May 2018, at 6:28pm, Simon Slavin wrote: >>>> >>>> I just realised that >>> >>> That was intended to be personal email. Apologies, everyone. >>> >>> Simon. >>> _______________________________________________ >>> sqlite-users mailing list >>> sqlite-users@mailinglists.sqlite.org >>> http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users >> _______________________________________________ >> sqlite-users mailing list >> sqlite-users@mailinglists.sqlite.org >> http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users >> _______________________________________________ >> sqlite-users mailing list >> sqlite-users@mailinglists.sqlite.org >> http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ > sqlite-users mailing list > sqlite-users@mailinglists.sqlite.org > http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ > sqlite-users mailing list > sqlite-users@mailinglists.sqlite.org > http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users