Any tips on how can I verify wich wall aligns with wich, and then group the aligned walls in lists? First I tried to retrieve each wall interior and exterior faces, then check for intersections, but it returned a list that I found too complicated do use. I have no clue how to cross-reference the intersections with the original walls.
This problem is more complicated then you would think at first.
How would you, for example, define walls that align and how could you then possibly group them? (From a coding perspective)
Yeah, I understand that’s quite tricky. From a coding perspective I do not have the slightest clue how to do it, since I can’t code.
I thought about some scenarios:
Group the walls by direction, then group again by paralellism, and at the end check if the bounding boxes intercept each other… But even without testing it, seems a little weary and uncertain.
Another scenario would check the distance between parallel wall centerlines, if they’re within some predetermined range, but it sounds even duller.
I was thinking of this like a first step to accomplish this: if I have a door or a window in any of these walls, Dynamo would recognize the sum of the grouped walls widths and adjust the door or window trim to fit the finished faces.
However, as it is getting complicated, I’m rethink my workflow begining from the door or window. I’ll work at this alternative and share the result if I succeed.
This is great! I was hopping for tips and you got the work done. Thank you so much!
If I may, running the dyn I’m getting an error at the group that creates smaller solids. I think that’s something about the -150 offset. The number is either too big or too small. May be the units? I’m testing in a model in meters.
There were 2 main reasons the script failed on your new/ more complicated walls:
The first bug was the way the biggest two curves were selected, I changed that now so that it always works, even when there are multiple windows/ doors in the wall and/or the wall has a non-rectangular shape (Haven’t tested it out on really weird shapes though).
The second problem was the occurence of one or more windows in a wall. When this occurs multiple correct perimeter curves (PolyCurves) can be extracted. This is now solved by grouping the curves first, then sorting them by length and then getting only the last item, which is always the PolyCurve representing the perimeter.
:
