Trimming points pulled from a function

Replies to This Discussion

Permalink Reply by Joseph Oster on March 11, 2017 at 4:50pm

You forgot to post your code.

Permalink Reply by Viktoria on March 11, 2017 at 4:55pm

Sorry for that! Here they are. 

Attachments:

• Coursework4_dc492_test3.gh, 407 KB
• GBR_Belfast.epw, 1.5 MB

Permalink Reply by Hyungsoo Kim on March 11, 2017 at 10:11pm

Hi.

I'm not sure this is what you are after. Anyway check attachment and reopen your epw weather file.

Attachments:

• Coursework4_dc492_test3_re.gh, 417 KB

Permalink Reply by Dan Cole on March 12, 2017 at 2:36am

Hi Hyungsoo, 

Viktoria posted this for me as I was waiting for my account to be approved, but this is exactly what I was looking for! Thank you so much for your time 

Permalink Reply by Hyungsoo Kim on March 12, 2017 at 7:36am

Fixed with BCX points per each line for proper data mathing.

Attachments:

• Coursework4_dc492_test3_reV2.gh, 417 KB

Permalink Reply by Joseph Oster on March 12, 2017 at 8:06am

I still don't believe that's correct, Hyungsoo?  There is a data tree mystery (GH bug?) using 'Graft' on 'Pull Point' that I wasn't able to solve until I came up with the 'Point In Brep (BrepInc)' idea.  This is your latest, in red, compared to mine in blue.  'BCX' points are red, 'Pull' points (with no grafting) are green:

This is your latest code (below), with my solution baked in blue:

See what I mean?

Permalink Reply by Hyungsoo Kim on March 12, 2017 at 9:02am

Hey, Joseph.
Well, frankly, in this case, I have little interest in what is right and wrong.
You methods may be right because the result is the connection of the lowermost points of the points on the surface with a curve.
If Dan wanted it, his setting from the first place  would not be necessary.
I just matched  each sun vector set to each column for pull points.

Permalink Reply by Joseph Oster on March 12, 2017 at 9:28am

Yeah, I tried MANY WAYS to do the same thing and failed.  Grafting the 'G' input to 'Pull' creates the error.  Without it, the puzzle here is that pull points in each tree branch are scattered among the 'ArrPolar' curves, yet collectively, they appear to be physically correct.  I tried 'Curve CP' on the pulled points to make the proper curve associations and that failed too.  Hence the extravagant approach of capped cylinders and 'BrepInc', which worked great.

Very weird.  Calling David Rutten?

Permalink Reply by Hyungsoo Kim on March 12, 2017 at 9:29am

For example, another approach...

Attachments:

• Coursework4_dc492_test3_reV3.gh, 427 KB

Permalink Reply by Joseph Oster on March 12, 2017 at 9:50am

Hah!  Yeah, skip the 'Pull' to lines altogether.  I considered that too, including use of 'Convex Hull', but didn't bother to pursue it.  Too much work, not my project.  ;)  Cheers

Permalink Reply by Joseph Oster on March 12, 2017 at 6:10am

I spent a lot of time on this last night...  Got results similar to (same as?) Hyungsoo Kim's solution, but believe it's not correct, based on a comparison to the pulled points and to the original 'BCX' points.

This morning, I took a different approach using 'BrepInc (Point in Brep)', to associate the pulled points with the 'ArrPolar' curves.  I created cylinders on those curves for 'BrepInc'.

This shows the difference between the two methods, mine in blue, Hyungsoo Kim's in red:

This is my solution compared to the 'BCX' points:

And the code:

Attachments:

• Coursework4_dc492_test3_2017Mar11a.gh, 418 KB

Permalink Reply by Joseph Oster on March 16, 2017 at 4:00pm

In a different thread, I created a 'PullT' cluster that does a much better job at associating pulled points with the geometry that pulls them.  It also applies to this thread.  The grafted pulled points are sorted by 'D (Distance)' and only the closest points for each curve are returned.

Attachments:

• Coursework4_dc492_test3_2017Mar16a.gh, 429 KB