I imagine that it was stuck and not detecting that it's stuck. Your new version worked (after I replaced g with exampleG on the last line. I'll try taking a look at it later today.
By the way, my email habits keep wanting me to change your definition of g to #: 51 435 636 652 398 445 154 947 761 121 (I know it's silly to impose email's line wrapping problems on an environment where they do not happen, but ... habit?) Thanks! -- Raul On Mon, Jul 9, 2012 at 12:14 PM, Max Harms <raeli...@gmail.com> wrote: > Sorry. I should've included actual data, rather than a random grid. Try > this: > > g =: 10 10 $ (0 0 0 0 1 1 0 0 1 1 0 1 1 0 1 1 0 0 1 1 1 0 0 1 1 1 1 1 0 0 1 > 0 1 0 0 0 1 1 0 0 0 1 1 0 0 0 1 1 1 0 0 1 1 0 1 1 1 1 0 1 0 0 1 0 0 1 1 0 1 > 0 1 1 1 0 1 1 0 0 1 1 1 0 1 1 1 1 1 0 0 1 0 0 0 1 1 1 1 0 0 1) > > Are you sure it's taking a long time, and not failing? It's likely that the > path from (5,0) to (0,8) on a random grid will fail. > > The graph search and hex grid search were hypothetical--i.e. I don't have > the path extension functions written. They were intended to be examples of > how aStar would be used in other contexts. > > The issues you brought up are probably correct. Yesterday I noticed those > lines were buggy and fixed them. Updated code is here: > http://pastebin.com/RjQA69j6 > > - Max > > On Mon, Jul 9, 2012 at 11:46 AM, Raul Miller <rauldmil...@gmail.com> wrote: > >> I'm running this right now... >> >> g =: ? 10 10 $ 2 >> 0 8 ((g gridExtendPath) aStar manhattanDistance) 5 0 >> >> ... but it's taking a long time. >> >> And without a working example that I can repeat, I cannot easily >> investigate the numbers. >> >> Meanwhile >> 'Z' (graphExtendPath aStar 1:) 'A' >> doesn't work because graphExtendPath was not in the pastebin. >> >> Still, this is enough to start trying to understand the code. >> >> Unless there's something going on that I do not understand (or unless >> there's something that *should be* going on that I do not understand), >> you should probably replace >> >> closed =: closed, last >> with >> closed =: ~. closed, last >> >> ...but, that said, the expression where closed gets used: >> if. +/ closed (*./^:_@:=)"_1 _ last do. continue. end. >> strikes me as odd. If closed is not empty and if 1 < # ~. last >> then this statement does nothing (because two different values in last >> cannot be equal to any single value in closed). Why so much work for >> edge cases? But another possibility is that the rank of closed was >> supposed to be 2 instead of 1. >> >> Anyways, perhaps you can cook up an example where it terminates >> quickly? That might make this easier to read. >> >> Thanks, >> >> -- >> Raul >> >> On Mon, Jul 9, 2012 at 10:14 AM, Max Harms <raeli...@gmail.com> wrote: >> > g =: ? 10 10 $ 2 >> > 0 8 ((g gridExtendPath) aStar manhattanDistance) 5 0 >> > >> > Hypothetical Dijkstra's Search on an arbitrary graph: >> > 'Z' (graphExtendPath aStar 1:) 'A' >> > >> > Hypothetical A* for hex grid: >> > goal (hexGridExtendPath aStar euclideanDistance) start >> > >> > - Max >> > >> > On Sun, Jul 8, 2012 at 11:02 PM, Raul Miller <rauldmil...@gmail.com> >> wrote: >> > >> >> On Sat, Jul 7, 2012 at 8:59 PM, Max Harms <raeli...@gmail.com> wrote: >> >> > Here's the functioning code if you're curious: >> >> http://pastebin.com/RjQA69j6 >> >> >> >> I'm curious. >> >> >> >> How do you use this? (What are some sample sentences.) >> >> >> >> [Especially when you are working with adverbs and conjunctions, but >> >> also even with verbs, examples help illustrate the relevant domains.] >> >> >> >> Thanks, >> >> >> >> -- >> >> Raul >> >> ---------------------------------------------------------------------- >> >> 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
