pairs.CSP {rflowcyt}R Documentation

Contour/Hexbin Scatterplot Matrices

Description

A pairs plotting of histograms and rectangular-binned or hexagonal-binned image plots are produced using hist and ContourScatterPlot, respectively.

Usage

pairs.CSP(x,
          status=NULL,
          box.idx.list=NULL,
          type.CSP=c("count.diff",
                  "p.hat",
                  "p.hat.norm",
                  "z.stat"),
          alternate.hexbinplot=FALSE,
          n.hexbins=100,
          range.x=range(x),
          varlabpos=round(seq(range.x[1],
                ceiling(diff(range.x)/150)*150+range.x[1],
                by=150),0),
          cutoffs = seq(range.x[1],
                  ceiling(diff(range.x)/25)*25+range.x[1],
                  by=25),  

           labels = colnames(x),

           panel = ContourScatterPlot,
           main="",
                    
           image.col=heat.colors(10),
           numlev=5,...,
           lower.panel = legend.CSP,
           upper.panel = panel,
           overlay.panel=rect.box.idx,
           border.boxes=1:length(box.idx.list),
           lwd.boxes=rep(3,length(box.idx.list)),
           lty.boxes=rep(1,length(box.idx.list)),
                      
           label.pos = 0.5,
           cex.labels = NULL,
           font.labels = 1,
           row1attop = TRUE,
           gap=1,
           ch.col=c("heat.colors(n)",
                   "rainbow(n, start=.7, end=.1)",
                   "terrain.colors(n)",
                   "topo.colors(n)",
                   "cm.colors(n)"))

Arguments

x matrix of data in which the columns are the variables and the rows are the individual observations
status numerical binary 0, 1 vector denoting the status of the observations; default is NULL
box.idx.list a list of vectors indicating the positions of 'x' which form a box to be overlayed on the binned plot in the upper and lower panels of the hexbin plot and the only the upper panel of the rectangular-binned plot by default
type.CSP character string denoting the type of value to be estimated using the 'status' for each cell grid: the difference in counts ("count.diff"), the proportion ("p.hat"), the normalized proportion at 0.5 ("p.hat.norm"), the z.statistic ("z.stat"), see make.density for details.
alternate.hexbinplot Boolean; if TRUE then alternate hexbin pairs plot is used; otherwise the ContourScatterPlot with rectangular bins is implemented
n.hexbins number of bins for hexbin call; default is 100
range.x vector denoting the min and the max of the observation values across all variable columns
varlabpos vector of position of the variable values in which to label the x and y axes
cutoffs the cutoffs for the x and y axes of the rectangular bins when alternate.hexbinplot is FALSE
labels the labels for the diagonals when alternative.hexbinplot is TRUE
panel default panel function; currently this is the contour scatter plot with rectangular bins; this option is ignored when 'alternate.hexbinplot' is TRUE
main the main title for the rectanglar Contour scatter plot when alternative.hexbinplot is FALSE
image.col image colors for the rectangular bins when alternative.hexbinplot is FALSE
numlev number of levels for the contours for the rectangular bins when alternative.hexbinplot is FALSE
... other options in hexagons or ContourScatterPlot
lower.panel function for the lower panels of the pairs plot; currently this is fixed as a hexbin (when 'alternate.hexbinplot' is TRUE) or the legend.CSP (when 'alternative.hexbinplot' is FALSE)
upper.panel function for the upper panels of the pairs plot; currently this is fixed as a hexbin or contour scatter plot
overlay.panel Function which describes the overlay image on the panels; currently this option only works with the 'rect.box.idx' function and other functions that have the same signature
border.boxes vector of corresponding border colors for each of the boxes in 'box.idx.list'
lwd.boxes vector of corresponding widths for each of the outlined boxes in 'box.idx.list'; default is for all the boxes to have lwd = 3
lty.boxes vector of corresponding line types for each of the outlined boxes in 'box.idx.list'; default is for all the boxes to have lty = 1
label.pos position of the labels on the diagonal panels which are currently fixed as histograms; this option is not in use currently.
cex.labels cex for the labels, used only when 'alternative.hexbinplot' is TRUE
font.labels font for the labels, used only when 'alternative.hexbinplot' is TRUE
row1attop boolean if row 1 is at the top, used only when 'alternative.hexbinplot' is TRUE
gap used only when 'alternative.hexbinplot' is TRUE
ch.col character string denoting the type of color palette used for the rectangular-binned image to be displayed in the legend when 'aternate.hexbinplot' is FALSE; default is "heat.colors(n)"

Details

There are no legends for the hexagonal (when 'alternate.hexbinplot' is TRUE) but there is a roughly estimate legend available for the rectangular binning (when 'alternate.hexbinplot' is FALSE) in the pairs plot.

Value

A pairs plot is displayed. NOTE: The histograms on the diagonals are of the whole dataset regardless of the value of the cells in each ContourScatterPlot.

Author(s)

J.Y. Wan and A.J. Rossini

References

Hexbin, other papers.

See Also

objects to See Also as 'hexbin' in the hexbin package

Examples

  if (interactive()){
    if (require(rfcdmin)){
      data.there<-is.element(c("st.1829", "unst.1829", "st.DRT",
                    "unst.DRT"),objects())

      if ( ( sum(data.there) != length(data.there) )){
         ## obtaining the FCS objects from VRC data
        data(VRCmin)
      }

      ## subsetting the data for quicker plot display of less data
      data.mat1<-st.DRT@data[1:10000, 1:5]

      ## hexagonal binning

      pairs.CSP(data.mat1, alternate.hexbinplot=TRUE)

      ## rectangular binning with legends 

      pairs.CSP(data.mat1, numlev=3,
         image.col=heat.colors(20))

      ## rectangular binning without legends 

      pairs.CSP(data.mat1, numlev=3,
         image.col=heat.colors(20),
         lower.panel=ContourScatterPlot)

      ## putting a box around the observations
      ## greater than 500 for the second variable
      ##  less than 200 for the first variable
      idx1<-which(data.mat1[,2] > 500)  ## green box
      idx2<-which(data.mat1[,1] < 200)  ## blue box

      box.idx.list<-list(idx1, idx2)
      ## hexbin plots
      pairs.CSP(data.mat1, box.idx.list=box.idx.list,
            alternate.hexbinplot=TRUE, border.vec=c("green", "blue"))
      ## rectangular binned plots
      pairs.CSP(data.mat1, box.idx.list=box.idx.list,
            alternate.hexbinplot=FALSE,border.vec=c("green", "blue"),
            lower.panel=ContourScatterPlot)
    }
  }

[Package rflowcyt version 1.14.0 Index]