Plotting Metrics Over Time

We can also examine how metrics for the overall graph change (or don't) through time. Earlier we looked at two important ones, clustering and reciprocity. Each were quite high, as we expected after visually inspecting the graph structure. However, over time, each of these might change. Are global purchasing patterns on Amazon stable? If we think so, then we expect plots of these metrics to essentially be horizontal lines, indicating that reciprocity is about the same every day and there's a high degree of clustering structure. Let's see what we can find here.

Code to calculate the transitivity by graph is shown.

This exercise is part of the course

Case Studies: Network Analysis in R

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

# Examine this code
transitivity_by_graph <- data.frame(
  date = d,
  metric = "transitivity",
  score = sapply(all_graphs, transitivity)
)

# Calculate reciprocity by graph
reciprocity_by_graph <- data.frame(
  date = ___,
  metric = ___,
  score = ___ )

Edit and Run Code

This exercise is part of the course

Case Studies: Network Analysis in R

BeginnerSkill Level

4.7+

Start Course for Free

In this chapter you'll explore a subset of an Amazon purchase graph. You'll build on what you've already learned, finding important products and discovering what drives purchases. You'll also examine how graphs can change through time by looking at the graph during different time periods.

Exercise 1: Exploring your data set Exercise 2: Finding Dyads and Triads Exercise 3: Clustering and Reciprocity Exercise 4: Important Products Exercise 5: What Makes an Important Product?Exercise 6: Exploring temporal structure Exercise 7: Metrics through time Exercise 8: Plotting Metrics Over Time

Current Exercise

In this lesson you'll explore some Twitter data about R by looking at conversations using '#rstats'. First you'll look at the raw data and think about how you want to build your graph. There's a number of ways to do this, and we'll cover two ways: retweets and mentions. You'll build those graphs and then compare them on a number of metrics.

Exercise 1: Creating your retweet graph Exercise 2: Visualize the graph Exercise 3: Visualize nodes by degree Exercise 4: What's the distribution of centrality?Exercise 5: Who is important in the conversation?Exercise 6: Plotting important vertices Exercise 7: Building a mentions graph Exercise 8: Comparing mention and retweet graph Exercise 9: Assortativity and reciprocity Exercise 10: Finding who is talking to whom Exercise 11: Finding communities Exercise 12: Comparing community algorithms Exercise 13: Visualizing the communities

In this chapter you will analyze data from a Chicago bike sharing network. We will build on the concepts already covered in the introductory course, and add a few new ones to handle graphs with weighted edges. You will also start with data in a slightly more raw form and cover how to build your graph up from a data source you might find.

Exercise 1: Creating our graph from raw data Exercise 2: Making Graphs of Different User Types Exercise 3: Compare Graphs of Different User Types Exercise 4: Compare graph distance vs. geographic distance Exercise 5: Compare Subscriber vs. Non-Subscriber Distances Exercise 6: Most Traveled To and From Stations Exercise 7: Most Traveled To and From Stations with Weights Exercise 8: Visualize central vertices Exercise 9: Weighted Measures of Centrality Exercise 10: Connectivity Exercise 11: Find the minimum cut 1 Exercise 12: Find the minimum cut 2 Exercise 13: Unweighted Clustering Randomizations Exercise 14: Weighted Clustering Randomizations

So far everything we've done has been using plotting from igraph. It provides many powerful ways to plot your graph data. However many people prefer interacting with other plotting frameworks like ggplot2, or even interactive frameworks like d3.js. In this lesson you'll look at other plotting libraries that build on the ggplot2 framework. You'll also look at other non-"hairball" type methods like hive plots, as well as building interactive and animated plots.

Exercise 1: Other packages for plotting graphs!Exercise 2: ggnet Basics Exercise 3: ggnetwork Basics Exercise 4: More ggnet Plotting Options Exercise 5: More ggnetwork Plotting Options Exercise 6: Interactive visualizations Exercise 7: Interactive plots with ggiraph Exercise 8: Interactive javascript plots Exercise 9: Alternative visualizations Exercise 10: Alternative ways to visualize a graph: Hive plots Exercise 11: BioFabric as an HTML widget Exercise 12: Plotting graphs on a map