fuse between groups only

[jwoelper]

Hi everyone,

just came across a problem where I wanted to fuse points between two or more groups.

Suppose I have groups A and B. I do not want points from group A to fuse with other points within group A, just with ones in group B. That way points in A should only fuse to ones in group B and vice versa.

Is that possible? I tried to work around that by creating volumes from the geometry and intersecting them, but that method is not precise and cpu-intensive...

Any help is greatly appreciated!

Johann

[dbukovec]

hi!

take a look a t the file, basically i created a new group based on other two, nothing magical

fuseGroups.hip

[jwoelper]

Hi!

Thank you very much for the help!

Unfortunately, this method does not work for me... my grouped points are closer together than their targets, so I'd strictly need a member of A snap to a member of B only.

I have included my problem next to your example in the file - sorry if I did not explain it properly.

fuseGroups.hipnc

hi!

take a look a t the file, basically i created a new group based on other two, nothing magical

[Mario Marengo]

Sounds like a perfect job for point clouds.

When I get a chance I'll sketch an example, but in the meantime:

 for each point 'i' in A
    open B as a pointcloud with query_pos=A[i] and radius=snap_threshold
       if point 'j' found, move A[i] to B[j]
    close pcloud
 end foreach

[Mario Marengo]

Here's a quick sketch using point clouds. I couldn't open your file Johann, so I modified Daniel's.

Just hit play to see it in action.

As an aside, while doing this I noticed that using pcfilter() (or using pcfarthest() as a determinant for whether a point was found, one or the other) failed in some cases (and shouldn't). No time to check right now, but it *may* be a bug. Anyway, you'll see that (much shorter/cleaner) approach commented out, and a hand-rolled pcfilter() used instead -- but still pretty straight forward.

fuseGroups_pcmethod.hip

HTH,

Cheers.

[mightcouldb1]

Thanks for taking the time to post this Mario. Super helpful as I am learning VEX.

[jwoelper]

Thanks a million for taking the time to assemble this example!

This was very very helpful!

Thanks again

Johann

Here's a quick sketch using point clouds. I couldn't open your file Johann, so I modified Daniel's.

Just hit play to see it in action.

As an aside, while doing this I noticed that using pcfilter() (or using pcfarthest() as a determinant for whether a point was found, one or the other) failed in some cases (and shouldn't). No time to check right now, but it *may* be a bug. Anyway, you'll see that (much shorter/cleaner) approach commented out, and a hand-rolled pcfilter() used instead -- but still pretty straight forward.

fuseGroups_pcmethod.hip

HTH,

Cheers.

[XAavBp]

Hi,

 

I have pretty much the same problem as you but i'm using a sweep to instance multiple objects along a curve, I've grouped all the points in the positive and negative  Z space separate group but I can't figure how to implement them with your point cloud.Could someone give me a hint as to how to implement your solution in my context.

 

Thanks !

 

 

test.hipnc