Question about the Buildings binary mask in SOLWEIG

[jsscmnhn]

Hi! I have a question about the building mask values in SOLWEIG when the land cover grid (lcgrid) is used for its creation.

In solweig_algorithm, it seems the building mask is generated by setting land cover class 2 to 0, and all other classes to 1, like this:

        buildings = cp.copy(lcgrid)
        buildings[buildings == 7] = 1
        buildings[buildings == 6] = 1
        buildings[buildings == 5] = 1
        buildings[buildings == 4] = 1
        buildings[buildings == 3] = 1
        buildings[buildings == 2] = 0

As I wanted to simplify this, and use cobblestone as class at 0, I changed this to:
buildings = (lcgrid != 2).astype(cp.float32)
This creates a binary mask where buildings = 0 and all other classes = 1.
My question is:
Is this logic correct? Or should it be the other way around; with buildings = 1 and everything else = 0?

I'm asking because in functions like gvf_2018a and sunonsurface_2018a, it seems more intuitive that buildings would be marked as 1 rather than 0.

Thanks in advance for the clarification!

[biglimp]

Buildings should be 0 and other surfaces should be 1. We agree that this a bit strange but (non-logic) but that how the code originally was set up so we have just kept it like this.

[jsscmnhn]

Thank you!