Homophily measures

library(spIBDerverse)

library(igraph)
#> 
#> Attaching package: 'igraph'
#> The following objects are masked from 'package:stats':
#> 
#>     decompose, spectrum
#> The following object is masked from 'package:base':
#> 
#>     union

In this vignette, we look at a method to identify edges that are more prominent than expected if the nodes attributes had not effect on the probability of an edge.

Consider a network with $N$ nodes and $E$ edges. From this, we get that the probability at two randomly selected nodes have an edge connected them is

$$p_{ij} = \frac{E}{{N \choose 2}}$$

obs <- example_network |> get_ringbauer_measures("site")
#> Warning: There was 1 warning in `dplyr::mutate()`.
#> ℹ In argument: `prop_test = purrr::map2(...)`.
#> Caused by warning in `stats::prop.test()`:
#> ! Chi-squared approximation may be incorrect