Dex: The Data Explorer


Dex is an application for interacting with data. Dex provides data extraction, transformation, visualization, and predictive capabilities.

Major capabilities include:

Data Extraction Dex has the ability to read data from a variety of sources such as CSV, files, databases and even processes and expect driven scripts.
Transformation Dex can transform this data into other forms via Groovy, Jython as was as a custom language specific to Dex called TMI (Too Much Information).
Data Persistence Dex can save the transformed data to file or database.
Visualization Dex provides 40+ visualization tools to help explore your data. In most cases the visualization can be saved into self-contained and interactive HTML5 content which can be viewed with or without a WebServer through a HTML5 compliant browser such as Firefox, Chrome, Safari or later versions of Internet Explorer. More visualizations are being added in each release.
R Integration Dex can integrate directly with a running instance of R to tap into complex statistical analysis and predictive analytics.


This list is by no means complete, but here is a short list folks who have helped Dex through inspiration and perspiration:

  • EJ-Technologies for donating a free copy of JProfiler. JProfiler is by far the best and most complete profiler available for Java.
  • Mike Bostock for his excellent JavaScript visualization library Data Driven Documents
  • Bob Monteverde for his D3 derivative work in NVD3.
  • Jason Davies for his many inspirational D3 based visualizations.
  • Jeffery Heer for his work in Prefuse. It is the root of much derivative works and thought around how we can better visualize data.
  • Bret Victor for his general brilliance and for opening my eyes to the concept of live coding and inventing on principle.
  • Mr. Doob is another source of inspiration. Unfortunately, I can't use WebGL in Dex yet, but his work is amazing and someday I hope that my work and his will coalesce.
  • The JavaFX guys John Smith, Jonathan Giles, Richard Blair and the Oracle team in general for having the guts to reinvent JavaFX after some missteps and the enthusiasm with which they support it.


Click on any of the images for a live version.

NVD3 Charts

NVD3 is an amazing set of charting capabilities which extends the capabilities of D3 to provide some very powerful visualization capabilities.

Sankey Diagrams

Great for proportion and flow.

Time Lapse or Motion Charts

Time lapse, or motion charts were inspired by Hans Rosling's Gapminder project and are used to depict complex relationships over time.

Groovy Templates

Groovy templates allow data transformation into any format. The following examples show a variety of transformations to HTML tables with differing capabilities such as highlighting and search capabilities. These views are not limited to HTML. Once generated, they can be saved as independent HTML files outside of Dex.

Basic Charting

All of the usual charting components are available. Many have multiple implementations with different capabilities.

Chord Diagrams

Chord diagrams are a visually stunning way to display relationships among a group of things.

Starburst and Partition Diagrams

Starburst and partition diagrams show hierarchical containment relationships.

Force Diagrams & Force Trees

Force diagrams and Force Trees are a fun way to show connections. They use physics to determine an optimal layout for a large number of nodes.

Parallel Lines and Radar Charts

Radar charts and Parallel Lines are great ways to depict relationships across a vast amount of numerical data.

SPLOM : ScatterPlot Matrices

SPLOM, or Scatterplot Matrices are very effective ways to compare numerical distributions of data.

Relational Views

Dendograms, hive plots, indented trees and node link trees are effectives ways to depict hierarchy and relationships.

3D Views

The Pre3D and Canvas3D task can be used to graph functions in three dimensions.

start.txt · Last modified: 2012/12/29 04:33 by admin
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki