Dijkstra goes outside of the country boarders for cell 41 of 100 and 41 is then nan km

[NMoksnes]

This leads to a nan result in the datafile. At the moment this is converted to 0.

[NMoksnes]

This was turned into a 0 to be able to get through the results. However, the optimal way to solve this is to limit Dijkstra to the country borders which might be conflicting with the padding that is done or the raster coarseness and the shapefile mismatch (when calculating the zonal statistics).