Open frederique-hub opened 9 months ago
Frederique developed notebook, this should be tested. Propose how to fix this.
Chamidu will:
If there is still time:
01) The visualization looks confusing because a location can be both isolated and flooded. Would be nice to to have column 'isolated/not isolated' , 'flooded/not flooded' 02) When taking out the edges that are flooded, sections containing bridges are not taken out (bridges can't be flooded). However if there’s a bridge in the middle of a fully/partly flooded edge the whole edge is taken as a bridge. When the bridges are not considered we get the desired output.
Might be useful to discuss the implications of taking out the bridges
The issue is explained here
For this, the example _example_isolatedlocations.ipynb can be used.
In this example the results look strange - there are locations that are isolated because of network disruption elsewhere but they seem to be disrupted due to the flooded road nearby. Also, there are some locations that are not isolated (in the South-East) but looking at the data, they should be.