Distance Mapping

Context & Objective

• The Distance Mapping layout algorithm visualizes measures of node similarity as distances.
• The algorithm's objective is to position similar nodes close together and dissimilar nodes far apart.
• Distance Mapping can use one of the following measures:
  • Mutual Information
  • Pearson's Correlation Coefficient R
• The Distance Mapping algorithm works with and without arcs, making it a useful tool for exploratory analysis before learning any network.

Usage

• Main Menu > View > Layout > Distance Mapping > Mutual Information.
• Main Menu > View > Layout > Distance Mapping > Pearson's Correlation.

Key Features:

• The layout algorithm uses an Expectation-Maximization algorithm and stops automatically when convergence is detected.

• If a node is selected when the algorithm starts, the automatic stopping feature is disabled. In this case, you must manually terminate the process by clicking the red status button in the Status Bar.
• While the layout process is running, you can select nodes and manually move them to influence positioning.
• Excluded nodes will not be affected by the layout algorithm.

Example & Workflow Animation

We use the voting records from the 112th U.S. Congress to illustrate the Distance Mapping layout algorithm:

• Each node represents a House member with three states: "yea", "nay", or "not voting/not a member".
• A total of 1,505 roll call votes were recorded for each representative in the 112th Congress.
• Nodes in blue represent Democrats, and nodes in red represent Republicans.