Talk:SegmentPlots.R

Could easily filter based on the "slope" of the arrow and color code those as well to get a feel for how much the M values change. Perhaps even use a color gradient linked to slope - horizontal is green, and then mix in blue (or whatever) as the slope increases. If you plotted the arrows in order of increasing slope the "bad" spots would stand out more clearly (i.e., they wouldn't be covered by the good spots). --Sven 11:54, 17 Oct 2006 (NZDT)

---

The concept here is to map spots which undergo two different transformations by arrows. Filtering of arrows can be made using different criterion, using heat map style colours the arrows can be plotted in a specific order to identify structure in the plots --Sven 17:02, 17 Oct 2006 (NZDT)

Input data could be organised in such a way as to take in two arguements which are the matrices you want to map with arrows. For each matrice element (e.g. X_ij → Y_ij, where i=i, j=j) is mapped to form the arrow. This allows fast prototyping of code for different normalization scenarios --Sven 13:48, 18 Oct 2006 (NZDT)