Luka Potkonjak has contributed a new feature to NodeXL: edge bundling. Following on the heels of the release of curved edges in version. 175, we now provide a way to gather parallel edges together into “bundles” that travel together until they diverge close to their destinations.
The effect is visually striking. Here is a before and after with the same network:
Curved Edges, no Edge Bundles
Curved and bundled edges.
In future releases we will expose more controls for configuring the bundles.
Recently, we’ve been setting out to set up a new way to set the option settings in NodeXL. Now we have added a new feature to NodeXL (v.166): savable, selectable configuration option settings files.
This may seem dull but this feature may have a big impact on the ease of use of NodeXL. This may let the most experienced users of NodeXL share their best practices with the rest of the user community.
Throughout NodeXL you can set a wide range of values, options, and settings. Change the default size of a font in the Graph Options and we record that setting. Map the size of a vertex to a value associated with it in Autofill columns and we record the setting.
Until now, we recorded the settings in several places: some settings went with each workbook, some settings were associated with each machine and stayed on each desktop. The result was that I could create a great network visualization, save and send someone the workbook and they often would not see what I saw when they opened and visualized the network graph. Why? People set their settings in lots of different ways, on different machines, creating potentially vastly different results.
Now, NodeXL will record everything about how a network workbook should be rendered in that workbook.
This now also means that expert users can save their settings for NodeXL and exchange them with other users.
NodeXL now has three places to put settings that should make it more reliable to share a workbook and get the same results on different systems. There will now be a default NodeXL settings file, a per NodeXL workbook settings file (stored in a hidden worksheet in the workbook) and any number of saved settings files. Users can save their settings in each workbook or decide to save the workbook’s custom set of settings to overwrite the default settings file that will be applied to all subsequent NodeXL workbooks. Users can also save their settings to one or more named files, which can be shared with others. Users can import any of these settings files and apply them to an open workbook by selecting NodeXL>Options>Import.
For those who are running automated collections that then run automated processing of a workbook (see: How to schedule the creation of a network with NodeXL and Windows Task Scheduler) you should note that:
Clusters are now groups in NodeXL. Recently, the NodeXL team has been focused on a set of new features related to grouping sets of vertices together. In the previous version we released a feature that allowed all sorts of groupings to be recorded in the worksheet. What’s new is that the three clustering algorithms we have already provided are just one form of group, components (connected sets of vertices) are another, and user labeled sets are a third method of creating a group of nodes in NodeXL (this last feature is still pending). This release adds the ability to add vertices to a group and then collapse all of the vertices in that group to a metanode – a composite of all the nodes in that group. It is then possible to expand the collapsed vertices into the graph
Here are the most recent features: 1.0.1.132 (2010-09-06)
1.0.1.131 (2010-08-19)
The NodeXL team has just released a new version (v.1.0.1.128) that contains a new “Automation” feature that allows users to define a collection of operations to perform on their network graphs and invoke the complete set in a single button click AND reuse that configuration on other workbook graphs. In fact, the feature will apply the configuration you define to all the files you specify, allowing easy processing of large collections of network data sets.
This week the feature is partially complete. Users can invoke the merge duplicate edges, calculate graph metrics, auto-fill columns, create sub-graph images, find clusters and show graph. These operations can require as many as dozens of clicks when performed manually. If you have dozens or hundreds of network data sets the result is a daunting case of repetitive strain injury and carpal tunnel syndrome. Instead, with automation, these operations can be carried out orders of magnitude more frequently without much pain!
The next release will feature the complete package which will then include control over the layout and graph options. As a result, automatically generated network visualizations can be produced in a pipeline: users will be able to specify a query using the NodeXL desktop network data collector and then automate the processing of large collections of data sets.
The result should be better analysis of time series data sets that have many “slices”. The feature points the way to additional development work for supporting the comparison between networks to evaluate their evolution.
The REM album “Automatic for the people” takes its title from the motto of Athens, Georgia, eatery Weaver D’s Delicious Fine Foods.
+
The Stanford Network Analysis Platform (SNAP) (http://snap.stanford.edu) is a high performance library for calculating network metrics of potentially very large graphs. Working with SNAP author and Stanford Computer Science Professor Jure Leskovec, the NodeXL team is releasing a new update with expanded support for network metrics. With SNAP integrated into NodeXL we have improved the scale and speed performance significantly (*very* significantly!). As of this release (v.1.0.1.122) the Betweenness Centrality, Closeness Centrality, and Eigenvector Centrality measures are calculated using the SNAP library. In addition we have added the Page Rank metric calculated by SNAP to the list of supported network measures. Two additional clustering algorithms automatically group nodes together into collections. With SNAP integrated into NodeXL we have added two clustering algorithms: Girvan-Newman and Clauset-Newman-Moore.
What network metrics matter most to you?
NodeXL updated starting with version 1.05 with features that make it fairly easy to create basic “Venn Diagrams”. A Venn diagram is a familiar way to illustrate the overlap (or lack thereof) of two or more “sets” of things.
There are some very amusing Venn diagrams out there! This one in particular made me laugh but I may be dating myself.
The Venn diagram feature is a special request from the Microsoft Biological Foundation group.
A Venn is related to but different from an Euler diagram. An “n-Venn” diagram is a collection of closed curves (“circles”) on a plane where all the circles intersect. A “simple” Venn diagram has just two circles but complex diagrams can have more. A 2 circle Venn diagram has 3 regions (A, B, A+B) and a 3 circle Venn diagram has 7 regions (A, B, C, AB, AC, BC, ABC).
A Survey of Venn Diagrams can be found at http://www.combinatorics.org/Surveys/ds5/VennEJC.html.
Our implementation is a bit of a hack, we basically let you define the X/Y location of 3 circles. A richer Venn tool would make it easy to take set data and define these circles. We may get that implemented in the coming months.
Version 95 of NodeXL is hot off the compiler and we are pleased to announce several major features that create a social media network analysis dashboard. From the NodeXL interface it is now possible to import networks from twitter, flickr, email, and a range of social network file formats. Coming soon: support for more spigots – the connectors that pull data from leading social media sources.
What social media data most interest you? We are considering integration with web and wiki crawlers, and support for YouTube, delicious, and enterprise data sources like Active Directory (LDAP), SharePoint, and Exchange.
This release also improves support for images, particularly those pulled from URLS, like twitter or facebook profile photos!
Here, for example, is a map of the connections among twitter accounts that tweeted the “WIN09” tag that was used in the recent Social Networks Summit at NYU (http://winworkshop.net/) The map illustrates the way the summit brought together previously separate clusters of people from the various disciplines that have been attracted to the study of networks in general and social networks in particular. Size of the image equals the number of tweets that person created.
A refined version adds Edge Labels and color to highlight the different tie types in the graph: “follows” relationships and “replies to” and “mentions” and now scaled by “Followers”.
In both views, the high betweenness role of one twitter account is clear.
Release details below the fold….