Testing parallelity of lines in x,y,z space and

Hi I've been stuck on this problem for a few days.

Is there a simple way to group a set of lines according to their parallelity?

Thanks,

James

Replies to This Discussion

Permalink Reply by Jesus Galvez on November 25, 2013 at 5:43am

Hello James,

you have to use the sort along curve and sort points components. With the first you can group parallel vectors , and with the other one you can group parallel lines.

Haven't tested it in 3D.

Attachments:

Group parallel.gh, 16 KB

Permalink Reply by James Park on November 25, 2013 at 6:05am

Jesus,

Thank you very much for you quick reply - your solutions are very nice. Can you think of a solution that would work for x,y,z space?

Thanks,

James

Permalink Reply by Jesus Galvez on November 25, 2013 at 7:13am

If you want it to work in 3D I suggest you used dot-product and/or cross-product operations.

Permalink Reply by Nick Tyrer on November 25, 2013 at 6:05am

Jesus, are lines only perpendicular if they are in same direction? wouldn't vectors that are reverse of each other also be perpendicular?

Permalink Reply by Nick Tyrer on November 25, 2013 at 6:27am

also the partition list doesnt work as its just roughly grouping the lines,

e.g. If there was 2x(0,10,0) and 8x(10,0,0) vectors, the partition list would just put 5 in each group irrespective of the perpendicularity. Sorry to be a buzz kill

I think i have a quick-fix for the first solution,

Permalink Reply by James Park on November 25, 2013 at 6:33am

Nick the grouping isn't really an issue, it's just arbitrary for graphical purposes (if I understand correctly). You can always create a set which finds all the distinct elements in a group and then do a member index search to find the respective index of each distinct element within the original set.

Permalink Reply by Jesus Galvez on November 25, 2013 at 7:01am

Don't worry for the "buzz kill", I also wanna learn new things!

Permalink Reply by Jesus Galvez on November 25, 2013 at 6:37am

Hello Nick,

I don't understand your question. Did you mean parallel as opposed to perpendicular? If so, that is why I did two versions, one for lines and another for vectors. Vectors are grouped if they point to the same direction; lines are group regardless of start-end direction.

Permalink Reply by James Park on November 25, 2013 at 6:49am

Jesus and Nick,

What I've essentially got is a set of lines generated by proximity. I can group them according to their vectors, i.e. direction simply by turning them into vectors, unitizing them and then making a set.

However, would be great if I could have set of lines grouped even if their vectors were in opposite directions. Please see attached file:

Thanks for your help so far :)

Attachments:

Groupparallel_reply.gh, 13 KB

Permalink Reply by Nick Tyrer on November 25, 2013 at 7:42am

James, here you go. There was alot of duplicate lines, so you may want to generate your structure differently.

Essentially i have sorted all the lines so that they attempt to point in the same direction. Then grouped them according to vector.

Attachments:

Groupparallel_reply v2.gh, 13 KB

Permalink Reply by Jesus Galvez on November 25, 2013 at 8:00am

Very nice. It seems it works only if the lines are parallel. There is no grouping if they are not completely parallel.

Here is one that groups vectors according to which way they point at. Note however that they are groups are grafted (one group for every vector). Hence if you want to create a singular set (instead of as many sets as vectors), you'll have to work it out some more.

Permalink Reply by Nick Tyrer on November 25, 2013 at 8:28am

cool. team work. I will think on it.