Display a data frame of annotation information

Description

Displays a data frame of both factor and numerical variables in parallel panels. Factors levels are indicated by black rectangles, using dummy variables for more than two levels. Numerical variables are shown as simple index plots with an optional loess smoother. Panels can be arranged horizontally or vertically, and different groups of subjects can be indicated through different background colors.

Usage

1
picketPlot(x, grp = NULL, grpcol, grplabel = NULL, horizontal = TRUE, asIs = FALSE, control = list())

Arguments

x

usually a data frame, which is passed to convAnnData to be converted to a numerical matrix with dummy coding for the factor levels. Alternatively, such a numerical matrix can be constructed manually and passed in as x, see argument asIs

grp

an optional vector of cluster memberships, in the same order as the rows of x

grpcol

an optional vector of background colors for the clusters specified in grp

grplabel

an optional vector of names for the clusters specified in grp

horizontal

logical value whether to plot variables horizontally (default) or vertically

asIs

a logical value indicating whether x should be passed to convAnnData for pre-processing or not. Defaults to FALSE.

control

a list of control parameters that determines the appearance of the plot

  • boxw is the relative length of the short side of a box marking (width for a horizontal plot). Default is 1.

  • boxh is the relative length of the long side of a box marking (default: 4)

  • hbuff is the relative distance between two box markings for the same variable (horizontal buffer for a horizontal plot). Default is 0.1

  • vbuff is the relative distance between two box markings for the same subject, but different variables (default: 0.1)

  • cex.label is the expansion factor for plotting cluster labels

  • numfac is the expansion factor indicating how much higher (for a horizontal plot) or wider (for a vertical plot) panels with numerical variables are than panels for factor variables.

  • nacol is the color for box markings indicating missing values (default: gray(0.85))

  • span is the span argument for the loess smoother. Default is 1/3; setting this to zero switches off smoothing.

  • degree is the degree of loess smoothing. Default is 1; setting this to zero switches off smoothing

  • pch is the plotting character for numerical variables

  • cex.pch is the size of the plotting character for numerical variables

  • col.pch is the color of the plotting character for numerical variables

Details

Missing values are indicated by a box marking in nacol for factor values.

Value

Depending on data and arguments, the value of the last axis function call. Irrelevant, as this function is called for its side effect of producing a plot.

See Also

annHeatmap2, convAnnData, par

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
    ## Standard call
    data(mtcars)
    picketPlot(mtcars)
    
    ## Pre-process the data for display
    mm = convAnnData(mtcars, inclRef=FALSE)
    picketPlot(mm, asIs=TRUE)
    
    ## Higher panels for continous traits
    picketPlot(mm, asIs=TRUE, control=list(numfac=3))
 
    ## With clusters
    picketPlot(mtcars, grp = rep(1:2, c(16, 16)), grpcol = c("pink","lightblue"), grplabel=c("Cluster 1", "Cluster 2"))