#TODO-List
This is more a scratch page to collect some ideas on what to do. It is not the definitive list of issues to fix.
As our intensities are only applied locally (and not in the full spatial coverage of our exposure model), we can skip the damage computation for a lot of cells (because lognorm cdf computation is expensive).
This is true especially for tsunami intensities as those only effects the coast line (and barely any cell behind).
However, we have to keep in mind that our fragility functions care about different units for the intensities, depending on the taxonomy.
It should be relativly easy to support grd files as inputs for intensity measurements.
At the moment the handling of the different columns is a bit inconsistent. We update the expo column (containing data from older runs), but we write new loss columns all the time (containing no data from older runs).
So it would be meaningful starting to aggregate the overall loss over mulitiple runs.