Finding median vector + grouping with a rule

Hi everyone,

Im doing a solar oriented geometry, that tries to maximize exposition. For explainability and simplicity i've only included 12 vector in this definition, but it is supposed to work with large sets of data (~3months, with 1 vector per hour). I have the solar vectors and i would like to add them on a given point, but only if:

the lines connecting the sun position to the center of the model do not intersect any given geometry. That is, i need to select from the group of lines, only the ones that do not intersect (BBX+CLX components) the brep. How can i do that?

Given this rule i also dont see how can i add several vectors, since i've only seen 2 vectors adding via the mathematical addition tool. Do i need to set them to one point?

[edit] i've solved this issue using mass addition tool for the 3 components (x,y,z) and feeding it with the flattened vector list.

Thanks in advance for any suggestion

Attachments:

definition.gh, 66 KB

Replies to This Discussion

Permalink Reply by Ethan on January 17, 2015 at 7:08pm

:) looking forward to seeing your results. hope this helps! -ethan

Attachments:

solar intersection.gh, 19 KB

Permalink Reply by Afonso Portela on January 18, 2015 at 8:08am

thank you very much, such a clever definition : ) Ill keep updating the results

Permalink Reply by Chris Hanley on January 19, 2015 at 12:29pm

Mesh|Ray intersect will do this for you.

The output H indicates hit or miss. You can then cull the list of sun vectors that intersect a building. Inverting the P input of the cull pattern component will give you the vectors that don't intersect.

(you can convert all of your building geometry to meshes, then join all the meshes). Even though the buildings are "separate objects", by joining them, GH will see them as one mesh, and calculate intersections accordingly.

In the example below, I only grabbed 2 of the building objects for simplicity purposes.

Attachments:

definition2.gh, 67 KB

Permalink Reply by Afonso Portela on January 19, 2015 at 12:37pm

Thank you Chris, i had already addopted ethan's method, but that seems really simple as well. Will try them both out to see which handles the most data, since i have almost 8 000 vectors to test.

RSS