4.1.2. Data Analysis

If you followed the example from the Data Input section of the companion website you should then be presented with the following image:

Note that all different example languages have been connected with each other, even language_d which had one "?" in its data.

Important: The most important fact when looking at and / or analyzing NeighborNet data is that the spatial distance between two endpoints is not important. What is crucial is the length of the shortest way (along the lines) from one endpoint to the other. While this makes no difference in this small diagram, the importance of this fact increases with larger projects (and thus larger, more complex diagrams).

The diagram above tells us these things:

- language_e and language_f are relatively similar

- language_c and language_a are relatively similar

- language_d and language_b lie somewhere in the middle

In order to read off / compare the distances between different items follow these steps:

Step 1

Open the distances tab on the left hand side of your screen. The window should look similar to this:

Step 2

You are now presented with the distances of the different items in matrix form. Each row represents the distance from one endpoint to all the other endpoints. For example:

The numbers in the first column, 0.0; 0.6; 0.2; 0.75; 0.8; .0.8; have the following meaning:

The distance from language_a to language_a is 0.

The distance from language_a to language_b is 0.6

The distance from language_a to language_c is 0.2

The distance from language_a to language_d is 0.75

The distance from language_a to language_e is 0.8

The distance from language_a to language_f is 0.8

Note: reading a matrix like this gets increasingly harder, the more different items you are comparing. Please consult the chapter on Complex Data - Data Analysis for a method to export this file into a more readable format.

Created with the Personal Edition of HelpNDoc: News and information about help authoring tools and software