About the separate configuration

In the README there is the following line:

"In the fourth phase, I used my own [ngraph.forcelayout](https://github.com/anvaka/ngraph.forcelayout) to compute layouts of nodes inside clusters, **_and a separate configuration to get the global layout of clusters_**."

How is the global layout of clusters calculated?

Start with the quoted passage in README and inspect the linked ngraph.forcelayout project to determine how the global layout of clusters is calculated. Done means documenting that calculation clearly in the README, including how the separate configuration relates to the cluster layout.

Written by the indexing model from the issue text.