When paths are linked to the hexagons that user's click on, they have to be down selected from the total number of paths used in the graph database grid cells. Furthermore, they have to be joined with the exit point to destination. This process is handled by this Juypter notebook. . Right now one path to each destination is chosen per hexagon; the shortest path to each destination type is the one chosen.
Looking at some polygons, there can be a more logical pathway for people to use in one part of a hexagon than another. For instance in the below screenshot provided by @etiennemutebutsi all pathways leading to schools, health services and markets are from the same side of the rivers. Services and pathways are also available on the other side of the river which would mislead the representation of the travel impact.
One potential fix suggested by @Mwalimu-Jimi is that we could use less granular paths and it just showed the general direction people go with and without bridges. The path would then just be the hexagons lighting up along a path. This likely would not be easier, but could be a more honest representation of the level of precision we are working with when estimating paths.
When paths are linked to the hexagons that user's click on, they have to be down selected from the total number of paths used in the graph database grid cells. Furthermore, they have to be joined with the exit point to destination. This process is handled by this Juypter notebook. . Right now one path to each destination is chosen per hexagon; the shortest path to each destination type is the one chosen.
Looking at some polygons, there can be a more logical pathway for people to use in one part of a hexagon than another. For instance in the below screenshot provided by @etiennemutebutsi all pathways leading to schools, health services and markets are from the same side of the rivers. Services and pathways are also available on the other side of the river which would mislead the representation of the travel impact.
One potential fix suggested by @Mwalimu-Jimi is that we could use less granular paths and it just showed the general direction people go with and without bridges. The path would then just be the hexagons lighting up along a path. This likely would not be easier, but could be a more honest representation of the level of precision we are working with when estimating paths.