Intro

Often links have associated weights. They may have two or more weights that we want to compare - for example:

• weights may be time-varying, and we may want to compare the values at two times
• we may have pairs of weights between pairs of nodes, corresponding to weights in each direction (for example, individuals may rate how friendly they are with others; more vehicles may travel along a segment of road in one direction than the other; more people may migrate from country A to country B than vice versa)
• we may want to compare weights from two different networks, given some correspondence between some nodes in one and some nodes in the other

The dataset in these examples is Newcomb Fraternity dataset, downloaded from CAOS and converted to JSON by a short Python script.

These 15 matrices record weekly sociometric preference rankings from 17 men attending the University of Michigan in the fall of 1956; data from week 9 are missing A "1" indicates first preference, and no ties were allowed.

The first two visualizations in this sequence of examples shows the ratings from each week; the remainder focus on the ratings in the first week.