Pure non-mathematical(sorry) solution to the increase in cluster & therefore load-time

[keepitopen]

1. For patient view: There are many non-relational nodes. How about combining these into super-nodes. For eg. 1 super-node represents 20 nodes. All individuals with no connections. Even if relational nodes increase these super-nodes can be configured in such a way that they contain only internal connections and not outwards. Nodes which cannot be combined this way are left as it is.

2. For City View(Maybe in Future): As covid cases have diversified in various cities, load times for city-view has increased. A non-mathematical solution to it could be to show the cities only state-wise and have outward directing arrows going to a new type of node(which represents inter-state connection). When person clicks on that node a smaller graph could open on the side. (Even better if we could this by hovering over the new type of node). These new directing arrows could be coloured differently to point that these point to patients in cities from another state. This will allow people to view the inter-state spread.

1 and 2 can also be combined in the future for state view, in case the need arises.

[someshkar]

This seems like an interesting solution, however I think for now a simple checkbox which removes all non-connected(nodes without edges) from the cluster works? @lovewithmind has made a feature like that in a PR already.