Finding Dyads and Triads
Let's think a little bit more about what we can learn from dyad and triad censuses of the overall graph. Because we are interested in understanding how items are purchased together and whether or not they are reciprocally purchased, dyad and triad censuses can provide a useful initial look. A dyad census will tell us how many items are purchased reciprocally vs. asymmetrically. The triad census will tell us which items might be important, specifically patterns 4, 9, and 12. All of these have one vertex that has 2 out degree edges and no in degree edges. Edge density should also give some insight we expect for graph clustering.
The Amazon co-purchase graph, amzn_g, is available.
This exercise is part of the course
Case Studies: Network Analysis in R
Exercise instructions
Explore amzn_g.
- Do a dyad census using the
dyad_census()function. - Do a triad census using the
triad_census()function. - Find the edge density using the
edge_density()function.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# The graph, amzn_g, is available
amzn_g
# Perform dyad census
___
# Perform triad census
___
# Find the edge density
___