hi all ,
i have 3 curves to join its not working
how to fix the internalized curve and then join
regards
rajeev
Curve issue.gh (11.7 KB)
If you bake your curves and explode them - you’ll see that each one is a closed curve made up of two duplicates - that’s why they aren’t joining
your curves arent joined. They need to be joined to each other in order to achieve one complete joined curve.
1 Like
No. They are three segments of which two are duplicates and the other is a closed curve. The closed curve is definately strange. It turns back onto itself, but does not give a discontintinuity at its midpoint…
If you pull apart the control points of the closed curve that can’t be exploded, you’ll see it’s also looping back on itself
Yes. I know. And said as much. It won’t shatter when normalized at 0.5, but you can take a subcurve from 0 to 0.5… don’t know how comfortable I am with that though.
@rajeev_pulari a solution, if you're comfortable with it?
Curve issue VR 01.gh (23.3 KB)
In all likelhood, this was a curve that originally turned back on itself, was split into three, and then already attempted once to be joined:
If there is a possibility to try to fix the original curve, I’d sooner do that.
1 Like
yes its closed curve not a duplicate one these curves are projected and then try to join
i tried with control points all curves have 3 points where the start and end point are the same
like a loop
so could not fins a method to break it and make a curve
Ah, yeah. That makes sense then.
Can you create a bounding box around the curves in the direction of projection and split them where they touch the box before projecting? Is the projection parallel? Or do you only have the resulting projection?
@rajeev_pulari Something like this?
Curve issue VR 02.gh (36.4 KB)
thank you
if the curves are non planar is there any method to do
regards
rajeev
For the planar case, you should be able to get the curves directly out of the Brep|Curve component in the script above, or if not distinguish between planar and non-planar cases and develop another solution, like intersecting the rectangle of the bounding box with the curve with Curve|Curve intersection.
If the curves have the same start and endpoint though, simply taking a subcurve of the normalized domain 0 To 0.5 should be fine…
Edit: Planar is horrendous determining if a curve is actually planar. Here, the segment that was an un-explodable closed curve, which is not planar, being detected as planar and the two resultant curves from Brep|Curve which are offset to each other:
Curve issue VR 03.gh (32.6 KB)
That puts the kaibosh on the whole thing.
Like I said: subcurve at 0.5
1 Like
i love it
thanks a lot