LB View Percent - Sky View Factor - error: "Input must be of type of Color"

Hello!

I am trying to do a Sky View Factor analysis, and I get a fatal error with this message:

1. Solution exception:Input must be of type of Color:
’NoneType’ object has no attribute 'r’

I know that the issue is with the input surface geometry, because I tested a simpler surface, and the component ran okay. Still, I don’t know how to fix the surface.
The input surface is a trimmed surface with cutout holes (building footprints). It is planar.


LB-view-percent-SkyView.gh (64.3 KB)

Does anyone know what this error means and why this surface is not accepted?

The grid size is set to 1 m. When I change to 2 m, the simulation runs fine. Is there a ceiling on the size of the analysis grid?

Cheers!
/Aga

@aguosha, maybe the grid size is too small. Check the size ratio of the surface, and start with large numbers.

1 Like

Hey @aguosha ,

@OmidmRashidi is correct but, more specifically, your grid size is very small compared to how far away you are from the Rhino origin. The distance between the Rhino origin and your scene is at least the distance between New York and Washington DC. On this scale, a meter is so small that it can’t really be represented with floating point numbers. So just move your scene close to the Rhino origin and it should be fine like you see here:


LB-view-percent-SkyView_CWM.gh (61.3 KB)

2 Likes

That is really good to remember. The model is often far from the scene origin when it is imported from geodata, which was the case here.

Thank you!!

I got the same error while working with a pretty large scene - rooftops as a single mesh for a whole city (just to test LB). I tried to load a southern part of rooftops, it was around 150 000 faces, it showed me this error. Then I tried a northern part (180 000 faces) that is located further from 0,0,0 and it worked (maybe around 3 minutes).
I suppose there is a degraded face or faces I can’t catch.
In the case of @aguosha it seems a degradation to happen due to the large coordinate values that impact the whole study mesh.

Breaking up the study into more manageable chunks is probably the best way to evaluate a whole city. This error results from the fact that the scene dimensions are so large in relation to the details of the geometry that they cannot be appropriately represented using the 15 digits available with doubles / floats. So you really have to either make the scene smaller or reduce it’s detail.

Thank you, @chris I did it just to test whether it can calculate it or not. In addition here is my post on breaking into chunks a study area Ladybug for large urban scenes, shading context and solar radiation
Any comments will be much appreciated!