(Classification approach) Last class gives always empty exposure

[gfdrr]

(Classification approach) Last class gives always empty exposure

[matamadio]

Actually maybe baby this is not a bad thing, allows to filter out over a threshold.
We can make it explicit, i.e. "last class includes value to be ignored".

[ConnectedSystems]

I'm seeing outputs for Class 6 (C6, the last class):

Or do you mean something else?

[matamadio]

This happens in the landslide notebook I tried to do, setting a smaller number of classes.

class_edges = OrderedDict({
    f'class_{i}': widgets.BoundedFloatText(
        value=k,
        min=0.0,
        max=10.0,
        step=0.0001,
        description=f'Class {i}:',
        tooltip=f'Minimum value of class {i}. Value must be less than the next entry.',
        disabled=True
    ) for (i,k) in zip(range(1,5), [0.0001, 0.001, 0.01, 0.05])
})

Am I doing it wrong? :(
Maybe safer if you stem out the class-only versions, and I will then edit parameters for each hazard:

1. Classes only, no RP loop - single hazard raster (for landslide and drought)
2. Classes only, with RP loop - just remove the "impact function" loop and choice (for heat stress and wildfires)

[matamadio]

Here is the bugged landslide notebook.

[ConnectedSystems]

Fixed in commit 957517d