An interesting problem

McGlinchy, Alistair Alistair.McGlinchy at
Wed Jan 3 20:40:58 GMT 2007

> On 3 Jan 2007, at 20:10, McGlinchy, Alistair wrote:
> > That's sort of what I have been playing with, but it still doesn't
> > seem
> > impressively fast. There's a large number of off-by-one errors  
> > possible
> > in here, but the answers seems to tally up when I check in Excel.
> Another thing you can do is have the quadtree (or whatever structure  
> you build) record the minima and maxima for each region. That'll  
> might allow you to discount some regions (too negative) without  
> having to sum them.

May need a but more clarification here. How would this work for the
following example. Consider a large matrix of -1 and +1 values with this
sub matrix in the middle

-1000  -1000  -1000  
-1000  +6000  -1000  
-1000  -1000  -1000  

That's -2000 as a 3x3 array. But you don't want to discount it as the
+6000 will likely be the best range alone.



Registered Office:
Marks and Spencer plc
Waterside House
35 North Wharf Road
W2 1NW

Registered No. 214436 in England and Wales.

Telephone (020) 7935 4422
Facsimile (020) 7487 2670


Please note that electronic mail may be monitored.

This e-mail is confidential. If you received it by mistake, please let us know and then delete it from your system; you should not copy, disclose, or distribute its contents to anyone nor act in reliance on this e-mail, as this is prohibited and may be unlawful.

More information about the mailing list