points accesibility test help ¡¡¡¡

For my PhD thesis I´m making a connectivity / accesibility test of a certain urban area streets.

i have a problem with the accesibility test, cause i need to know wich points of the grids have direct connections with the others, as the image example

Any idea, to how finally get a list with all the points and their direct accesibility correspondence?????

Thanks ¡¡¡¡¡¡¡¡¡¡¡

Ps: here are the 3dm & Gh files

PhD_Connectivity_Test.3dm

PhD_Connectivity_Test.gh

Replies to This Discussion

Permalink Reply by Mateusz Zwierzycki on January 3, 2013 at 1:56pm

You can use polylines tools from starling plugin. There are some topology analysis components, which will produce this kind of lists and even more.

Permalink Reply by Michael Pryor on January 3, 2013 at 2:10pm

As Mateusz says with his plug in. It changes the point order to match the curve order but its the same grouping of points just different label order.

Permalink Reply by Michael Pryor on January 3, 2013 at 3:03pm

I would also like to note that starling sorts the points in radial order around the source point which is a nice feature.

Permalink Reply by Hannes Löschke on January 3, 2013 at 2:03pm

Damn... took me a while because of one of those freaky wrong connections that freeze GH.

So here's my attempt..

Attachments:

PhD_Connectivity_Test_connectivity_tree.gh, 33 KB

Permalink Reply by newy011 on January 3, 2013 at 2:28pm

Has anyone managed to actually solve this yet or are these just solutions that work for some, not for others, or require some messy manual or patch-type workarounds to function correctly rather then solving the problem explicitly? Its difficult to tell from the images but most look like short-cuts or compromised?

Permalink Reply by Michael Pryor on January 3, 2013 at 2:31pm

they are working correctly,just all different approaches, One with script, one in gh only, the other 2 with plug ins. the starling one just re-orders the list, but its the same grouping. But I don't think it really matters if the point is called 0 or 1 or 2 ect. as long as its paired with the corresponding correct points.

Permalink Reply by Hannes Löschke on January 3, 2013 at 3:27pm

It's actually two different GH only approaches ;)

My approach is "limited" to a network of line segments that connect at the "intersections". But you could easily create such a network from any set of curves with MCX, Shatter and by connecting the start and end points of the segments with staight lines.

btw.: how can you tell from the posted images that tey are short-cuts or compromised?

Permalink Reply by Michael Pryor on January 3, 2013 at 3:58pm

well you know what I mean was just expressing that there are a lot of approaches :)

Permalink Reply by newy011 on January 3, 2013 at 4:07pm

Going back to my initial point where we as humans can work this out easily, but a computer cant. Just by looking at these solutions they don't really appear to following any kind of logic which suggested a work-around or manual fix had been applied post 'solution'.

The idea in itself seems fuzzy anyway, so in that regard maybe it is solved but also it 'isn't' solved! One of those paradoxical problems only human folly could ever produce. I'd rather be a rock.

Permalink Reply by Hannes Löschke on January 3, 2013 at 5:11pm

Jorge siad, he has a given steet map. This already is a kind of connectivity diagramm. It's just a graphical representation of it and he needs the topological info extracted.

So the question really is: given a set of lines connected at points, which are the points at the other end of any line directly connected to my current point? Or to use the steetmap point of view... Which other intersections can I reach without crossing an itersection?

...not very fuzzy to me, or am I missing some important part?

Permalink Reply by Michael Pryor on January 3, 2013 at 5:43pm

The point is not a shortest walk problem. The point is that this map is a pre defined street map. In which case the computer def solved the list of point to point categorization faster then I could write it down.

Permalink Reply by Michael Pryor on January 3, 2013 at 3:01pm

Here is how to tell if its correct as Djordge shows here. The point number will be the branch number in his case point 6 is branch(6) and in that branch will be the list of points that point touches based on the lines.