Pruning links using nearest neighbor algorithm
Networks with large numbers of edges (links between nodes) can be difficult to view and analyze. In such cases, applying Minimum Spanning Tree logic to the network might be helpful. In the context of a network, we call this the Nearest Neighbor method. If you select Nearest Neighbor under the Filtering tab ( see Figure), the algorithm prunes the network such that for each node, it retains only links that connect it to its closest neighbors, i.e. the links with the smallest distance. However, it allows the user the option to connect more than just the nearest neighbor. This can be done by specifying an epsilon value– a value that sets the limit for MST weight increase. The default value is set to 0.001. A slider bar allows you to increase the epsilon value, and you will see edges being added back as the value increases.
