Using Tensors to Analyze a Large Portfolio of Stocks
A large portfolio of stocks (more than 4800) is the raw data for this Demonstration.
The data is from the close of business on October 21, 2011. The data is mapped into an -dimensional tensor and then functions are evaluated for each of the respective nodes. The tensor is then mapped into two dimensions, and labels are generated and placed into a grid. Attributes based on percentages are displayed with the RedGreenSplit gradient to provide greater visual impact.
The overall experience of this Demonstration is similar to a typical pivot table within a spreadsheet, but given the overall richness of Mathematica, the capabilities are much broader and potentially complex. The code was designed to be as dynamic as possible and to make as few assumptions regarding the size of the data and the dimensions themselves.