Review PCA results using Tk graphics

The advantage of the Tk plots (available within the CompClust IPlot package) is that some are designed to be interactive. Users can click on data points or line plots to find out the name of an interesting vector (e.g. to get the identity of an extreme gene).

To create the interactive plots, first we need to create the IPlot-based PCAGinzu object. In the case of the yeast cell cycling dataset this is fairly quick, but can take tens of minutes to nearly an hour for something like a 33,000 x 158 dataset such as the GNF dataset.

ipcaginzu = IPlot.PCAGinzu(cho,outlierCutoff=0.05)


Joe Roden 2005-12-13