Figure out which algorithms are used for node placement in the graph visualization framework

[JonathanReeve]

It's probably somewhere in here, in vis-network

[JonathanReeve]

Tian writes:

In the discussion of layout
https://visjs.github.io/vis-network/docs/network/layout.html

Example:

"When enabled, the network will use the Kamada Kawai algorithm for initial
layout. For networks larger than 100 nodes, clustering will be performed
automatically to reduce the amount of nodes. This can greatly improve the
stabilization times. If the network is very interconnected (no or few leaf
nodes), this may not work and it will revert back to the old method.
Performance will be improved in the future."

Google "Kamada Kawai algorithm" gives

• https://en.wikipedia.org/wiki/Force-directed_graph_drawing (vague and high
  level explanation)
• Original Paper:
  https://www.sciencedirect.com/science/article/pii/0020019089901026
• A review:
  https://cs.brown.edu/people/rtamassi/gdhandbook/chapters/force-directed.pdf

Here is another place
https://visjs.github.io/vis-network/docs/network/physics.html

On this page, it mentioned

"You can select your own solver. Possible options: 'barnesHut', 'repulsion',
'hierarchicalRepulsion', 'forceAtlas2Based'. When setting the hierarchical
layout, the hierarchical repulsion solver is automatically selected,
regardless of what you fill in here."

It would important for us to understand these different options. Here are some
papers I found by googling.

• https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.357.6007&rep=rep1&type=pdf

• https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0098679