Contents - Index


Entry stratification

 

 

Entry stratification consists of a groups of functions to generate subsets of entries based on their placement in the biplot. The following options are available:

 

When this function is invoked, GGEbiplot first calculates the vector length of all Entries, and then ask the user to set a criterion, against which Entries with LONGER vectors are selected and retained. A new biplot based on the selected subset of Entries will be generated. Entries with longer vectors are more responsive to the (change of) Testers.

 

When this function is invoked, GGEbiplot first calculates the vector length of all Entries, and then ask the user to set a criterion, against which Entries with SHORTER vectors are selected and retained. A new biplot based on the selected subset of Entries will be generated. Entries with shorter vectors are less responsive to the (change of) Testers.

 

When this function is invoked, GGEbiplot first calculates the angles of the Entries, relative to an arbitrary reference point. It then asks the user to set a sector defined by an upper and a lower bound; all Entries WITHIN this sector will then be retained and a new biplot generated. 

 

When this function is invoked, GGEbiplot first calculates the angles of the Entries, relative to an arbitrary reference point. It then asks the user to set a sector defined by an upper and a lower bound; all Entries OUTSIDE this sector will then be retained and a new biplot generated.

 

When this function is invoked, GGEbiplot first identify entries that would be the vertex entries in the which-won-where view, and then generates a biplot based on these entries. The vertex entries are the most responsive entries.

 

When this function is invoked, GGEbiplot first calculates the means of the entries across testers and scaled by the largest mean of all entries . It then asks the user to set a criterion based on which entries with high means are selected. This function may not make sense when the entry main effect (G) is too small relative to the interaction (GE). When this is the case, there will be wide angles (i.e., negative correlations) between the Testers.

 

When this function is invoked, GGEbiplot first calculates the means of the entries across testers and scaled by the largest mean of all entries . It then asks the user to set a criterion based on which entries with low means are selected. This function may not make sense when the entry main effect (G) is too small relative to the interaction (GE). When this is the case, there will be wide angles (i.e., negative correlations) between the Testers.