The Sigma chart is a node graph created in d3 and meant to provide a helpful way of navigating through a hairball of relationships. The different features that help to detangle the hairball are the following:

###Features:

• Zoom by scrolling your mouse
• Pan by clicking and dragging
• Mouse over the node to show only connected nodes and dashed lines when not
• Relative node/text size/color based on measure proportionality
• Relative line width/color based on measure proportionality
• Click and drag a node to manually place them for desired positioning

###Dimension/Measure configs:

• Dimension1: Source Field (e.g. Product1)
• Dimension2: Target Field (e.g. Product2)
• Measure 1: Edge Count/Weight (e.g. sum(MutuallyOccurringFrequency) )
• Measure 2: Node Count/Weight (e.g. sum(Sales) )

###Shortcomings:

• Selections require mirrored pairs of data; for example, existence of Source and Target would require to have Target and Source as well.
• If selections are made and the other side of the pair is not there, it shouldn't error out, just disallow the selection

###TO DO:

• Make selection-related data availability more self-service friendly, ie, dynamically generate pairs if not available
• Cluster groups of nodes together
• Paths and Arrows pointing in most traversed direction
• Employ d3.linear() function for proportional node/edge styling
• If data points over a certain amount, map/reduce to higher grain and drill in, OR employ Fisheye

Any help on above ToDO's would be appreciated via Pull Request.

###Screenshot: