median vector of two vectors

Replies to This Discussion

Permalink Reply by Daniel Piker on September 22, 2013 at 5:23pm

I think what you are looking for is the average (mean):

(cross product is something else entirely)

Permalink Reply by Johannes on September 23, 2013 at 5:03pm

thank you for the tip but it does not work as I expected.
This tool gives me two vectors, in one point and in two directions!? and they have the same orientation like the two surface vectors which i loaded before.
Can you (or someone else) help me to find the average vector from two surface vectors in my exsample file?

Greetings,

Jo.

Attachments:

• help_find_average_vector.3dm, 73 KB
• gh_help_find_average_vector_01.gh, 6 KB

Permalink Reply by David Rutten on September 24, 2013 at 2:55am

One of your sliders is set to -10, meaning your line segment actually points in the opposite direction of the vector. The average vector is thus weird as both input vectors are almost entirely anti-parallel. I'd use a Vector Display component (perhaps with a V*10 expression in the V input if your vectors are too short to see):

Also note I flipped one of the surfaces so that the normal is pointing in the right direction.

--

David Rutten

david@mcneel.com

Tirol, Austria

Attachments:

• gh_help_find_average_vector_01.gh, 11 KB

Permalink Reply by David Rutten on September 24, 2013 at 2:57am

If you use [Interpolate Data] instead of [Average], then you can smoothly change the position and direction of the points and vectors using a slider:

--

David Rutten

david@mcneel.com

Tirol, Austria

Attachments:

• InterpolateVectors.gh, 11 KB

Permalink Reply by Johannes on September 24, 2013 at 3:10am

ah, thank you! i will try this later. Please ignor my thread below!

Permalink Reply by Johannes on September 24, 2013 at 9:05am

I wasnt able to try your definition but when I take a longer look to it, it seams to be that you create a vector which changes its direction depending to his position between the two other vectors, is that right? If yes, then I'm looking for something different. Namely I would like to extrude a new surface (on the bordered edge of both surfaces) which tooks the exactly the middle direction (half of the angle). Therefor i need the "summtion" of both vectors which has the same orientation in every point, independand from the distance to the two vectors.
Hope you can catch my meaning.
But i will check your definition anyway.

Permalink Reply by Johannes on September 25, 2013 at 6:14pm

Hello David,

i was wrong, your definition works quite good for me when I do this for only one pair of surface. Thank you for that.

The prolem was, and still is, the fliped surface. For a unknown reason, some of the surfaces in my whole geometry are fliped, even though they all were made within the same process.
Therefor I can not use the fliped tool here in generell, when i would like to serve more than one paire odf surfaces, because GH doesnt know what is right or wrong.
Is it possible to flip a bunch of surfaces with variable orientations into one equal direction?

Permalink Reply by David Rutten on September 26, 2013 at 2:23am

The only way to do that is to join them together into a single polysurface. Rhino doesn't expose the normal flipping logic on its own. Problem with joining/exploding them is that the order will get messed up. So you have some options:

• Join/Explode, then sort and redistribute the surfaces again. This may be easy, this may be impossibly difficult depending on how the surfaces were distributed.
• Fix the input by hand before you even get to Grasshopper. Could be the least amount of work, but it also means you cannot fully automate the process.
• Perform the flipping yourself based on some algorithm. Again, this could be easy or very difficult depending on how your surfaces are organised.

If you know all your surfaces must face away from a central point somewhere, then you can just evaluate the normal somewhere on the surface, figure out whether it pointing towards or away from a fixed point and then flip accordingly.

--

David Rutten

david@mcneel.com

Tirol, Austria

Permalink Reply by Johannes on September 26, 2013 at 12:21pm

ok. I have to try out which points fits to my requirements.
Thank You for your help once again!

Permalink Reply by Johannes on September 26, 2013 at 12:32pm

Wait, could it be made that way,  so that each surface in a list flips in the same direction (or not) as the surface before/above? doesnt sound that difficult. I will try what I can get..

Permalink Reply by Johannes on September 24, 2013 at 2:58am

..maybe somebody knows a much easier way to find the bisector surfaces/plane between two (bordering) surfaces/planes???