Hello,

Anyone know how to make use of the variadic argument “farthest” for the vex intersect function?

Taking for granted the parameters are set up correctly for the following and not using “farthest”, this will work:


But if I add “farthest” to the function it errors out:


I've tried using different variations like instead of the string, just put in 0 or 1 ( maybe it's a truth ), and even putting in a defined vector variable ( maybe it just needs to set a variable ).

Still can't seem to make use of it.

Any suggestions?

Hi,

it looks like that you can only use this function, if you put the full path name of the geometry into it, and you need another parameter (int) to set it to 1 (if set to 0 it will be the nearest intersection). There should be an overloaded function to make this easier … .
I haven't tried everything, but there should be a better solution.

Thanks Aizatulin,

Would have taken me forever to get that combination of ‘farthest’ and an int value.

On that point, I think it's worth an RFE to update and clarify the docs.

But the needing to use op: is odd since what this means is that when using ‘farthest’ it will only work if referencing an external node, since even op: defining the current node will not work.

Will have to mention that to support and see what they say.

“intersect_all” is another option, where you get an ordered point array from nearest to farthest point and where you don't have to use the ‘op..’ string input.

Yeah..thanks for the tip…I was trying that with my original application (intersect_all).

But I had the issue that I couldn't use 0 tolerance, only a value very small and close to 0.

Maybe I'm wrong and will have to re-visit; I might be mistaken in my approach one way or the other.

Yes, usually tolerance values close to zero should produce more accurate results, but the algorithm may probably take more time to converge. If the tolerance is set to zero, the algorithm may not converge unless it has other safety conditions to jump out, but I have no idea about implementation details. Usually 0.001 is not a bad value for tolerances (my personal opinion/experience)

BabaJ
Thanks Aizatulin,

Would have taken me forever to get that combination of ‘farthest’ and an int value.

On that point, I think it's worth an RFE to update and clarify the docs.

But the needing to use op: is odd since what this means is that when using ‘farthest’ it will only work if referencing an external node, since even op: defining the current node will not work.

Will have to mention that to support and see what they say.

//i@Found  = intersect(geoself(), point(0, "P", 0), Direction,  Intersection,  U, V); // Works 

//i@Found  = intersect(geoself(), point(0, "P", 0), Direction,  Intersection,  U, V, 'farthest', 0); // Does Not Work
//i@Found  = intersect(0, point(0, "P", 0), Direction,  Intersection,  U, V, 'farthest', 0); // Does Not Work

i@Found    = intersect('op:/obj/geo1/merge1', point(0, "P", 0), Direction,  Intersection,  U, V, 'farthest', 0); // Works. - 'merge1' is previous node to wrangle using intersect
i@Found    = intersect('op:/obj/geo1/PW_Intersections', point(0, "P", 0), Direction,  Intersection,  U, V, 'farthest', 0); // Does Not Work. - 'PW_Intersections' is node of wrangle using intersect function

Thanks, the ‘farthest’ argument is really confused me.

Hi

There is another working variant (in wrangle)

So, instead of explicitly write a path to a node, you can just use the wrangle's @OpInputX syntax.

you can also directly do this