How to show panels for factor levels of conditioning variables which do have no values? E.g. there are panels for "Grand Rapids" when they have values: data( barley ) with( barley, dotplot(variety ~ yield | year * site, layout=c(6,2) ) ) There are no panels for "Grand Rapids" when there are no values for "Grand Rapids": my.barley <- subset( barley, ! ( site == "Grand Rapids" ) ) with( my.barley, dotplot(variety ~ yield | year * site, layout=c(6,2) ) ) But there is a level "Grand Rapids": levels( my.barley$site ) [1] "Grand Rapids" "Duluth" [3] "University Farm" "Morris" [5] "Crookston" "Waseca" Is there an option to show empty panels for "Grand Rapids" in ``my.barley''? Thanks. Wolfram
lattice: showing panels for factor levels with no values
2 messages · wolfram at fischer-zim.ch, Deepayan Sarkar
On Thursday 12 February 2004 03:12, Wolfram Fischer - Z/I/M wrote:
How to show panels for factor levels of conditioning variables which do have no values? E.g. there are panels for "Grand Rapids" when they have values: data( barley ) with( barley, dotplot(variety ~ yield | year * site, layout=c(6,2) ) ) There are no panels for "Grand Rapids" when there are no values for "Grand Rapids": my.barley <- subset( barley, ! ( site == "Grand Rapids" ) ) with( my.barley, dotplot(variety ~ yield | year * site, layout=c(6,2) ) ) But there is a level "Grand Rapids": levels( my.barley$site ) [1] "Grand Rapids" "Duluth" [3] "University Farm" "Morris" [5] "Crookston" "Waseca" Is there an option to show empty panels for "Grand Rapids" in ``my.barley''?
No, unused factor levels are dropped. I'll see if adding an option for this is feasible. Deepayan