Eliminate less important markers from a QQE biplot
- Under menu VBV Biplot, click Eliminate Short-vectored Markers, an input box will appear, asking for the name of a marker.
- Markers (covariables, in general) with vectors shorter than that of the marker that is used as a criterion will be eliminated from the biplot. They can always be called back by clicking Data/Back to Previous Subset.
- The overall importance of a marker is measured by the length of its vector (the distance between the marker and the biplot origin). A short vector means the marker has relatively small effect on the trait in all or most environments.
- The relative vector length will change with the singular value partitioning method. The entry-metric partitioning should be used for eliminating markers. This is the default method for this function. Clicking Relation among Testers under Biplot Tools will also bring up the tester-metric scaling, and clicking Relation Among Entries will bring up the entry-focused scaling.