[BioC] FlowViz and FlowStats Interaction in Gates on densityplots

[Aric Gregson]

Hello,

I am again running through some older code with flowViz and flowStats in
R and finding new incompatibilities. I have also found some very nice
enhancements. :)

I have very many rangeGate objects from flowStats that I had been using
to plot gates on densityplots from flowViz. This worked like a charm
previously. Now however, it seems that densityplot complains when one of
the values is NA, NaN or Inf. Of course, rangeGate creates values that
often (always?) include 'Inf'. The workaround is to create a manual gate
that does not have 'Inf' as the higher end on the gate, but clearly that
negates the entire reason for using reproducible code and R. Example
code as below.

filter_range_exosome <- rangeGate(Data(wf[['rect_bead+']]),
                              stain="FL2.H",
                              filterId="Exosomes",
                              plot=FALSE,
                              sd=0.1
                              ) 

densityplot(NAME ~ ., 
            Data(wf[['rect_bead+']]),
            par.strip.text=list(cex=0.7),
            channels="FL2.H",
            main="Exosome Population",
            filter=filter_range_exosome,
            alpha=0.4,
            xlab="CD9-PE",
            groups=TYPE,
            fitGate=TRUE
            )

Produces the following notice in the plot:

'Error using packet 1 'to' cannot be Na, NaN or infinite.'

I'm am certainly missing something, perhaps a new function for
developing a rangeGate that works with densityplot? 

If needed I can create some reproducible example with the GvH data, but
I think this is clear enough to get it started. 

Thanks very much, Aric