plot_AbanicoPlot: Function to create an Abanico Plot.

Description

A plot is produced which allows comprehensive presentation of data precision and its dispersion around a central value as well as illustration of a kernel density estimate of the dose values.

Usage

plot_AbanicoPlot(data, na.exclude = TRUE, log.z = TRUE, central.value, 
    centrality = "mean.weighted", dispersion = "sd", plot.ratio = 0.75, 
    rotate = FALSE, mtext, summary, summary.pos, legend, legend.pos, 
    stats, rug = FALSE, kde = TRUE, hist = FALSE, dots = FALSE, 
    y.axis = TRUE, error.bars = FALSE, polygon.col, bar.col, 
    line, line.col, line.label, grid.col, bw = "SJ", output = FALSE, 
    ...)

Arguments

data

data.frame or RLum.Results object (required): for data.frame two columns: De (data[,1]) and De error (data[,2]). To plot

na.exclude

logical (with default): exclude NA values from the data set prior to any further operations.

log.z

logical (with default): Option to display the z-axis in logarithmic scale. Default is TRUE.

central.value

numeric: User-defined central value, primarily used for horizontal centering of the z-axis.

centrality

character or numeric (with default): measure of centrality, used for automatically centering the plot and drawing the central line. Can either be one o

dispersion

character (with default): measure of dispersion, used for drawing the polygon that depicts the dose distribution. One out of "sd" (standard deviation),"2sd" (2 standard deviati

plot.ratio

numeric: Relative space, given to the radial versus the cartesian plot part, deault is 0.75.

rotate

logical: Option to turn the plot by 90 degrees.

mtext

character: additional text below the plot title.

summary

character (optional): adds numerical output to the plot. Can be one or more out of: "n" (number of samples), "mean" (mean De value), "mean.weighted" (error-weighted mean), "median" (median of the De

summary.pos

numeric or character (with default): optional position coordinates or keyword (e.g. "topright") for the statistical summary. Alternativel

legend

character vector (optional): legend content to be added to the plot.

legend.pos

numeric or character (with default): optional position coordinates or keyword (e.g. "topright") for the legend to be plotted.

stats

character: additional labels of statistically important values in the plot. One or more out of the following: "min", "max", "median".

rug

logical: Option to add a rug to the KDE part, to indicate the location of individual values.

kde

logical: Option to add a KDE plot to the dispersion part, default is TRUE.

hist

logical: Option to add a histogram to the dispersion part. Only meaningful when not more than one data set is plotted.

dots

logical: Option to add a dot plot to the dispersion part. If number of dots exceeds space in the dispersion part, a square indicates this.

y.axis

logical: Option to hide y-axis labels. Useful for data with small scatter.

error.bars

logical: Option to show De-errors as error bars on De-points. Useful in combination with y.axis = FALSE, bar.col = "none".

polygon.col

character or numeric (with default): colour of the polygon showing the dose dispersion around the central value. To disable the polygon use "none"

bar.col

character or numeric (with default): colour of the bar showing the 2-sigma range of the dose error around the central value. To disable the bar use

line

numeric: numeric values of the additional lines to be added.

line.col

character or numeric: colour of the additional lines.

line.label

character: labels for the additional lines.

grid.col

character or numeric (with default): colour of the grid lines (originating at [0,0] and strechting to the z-scale). To disable grid lines use "n

character (with default): bin-width for KDE, choose a numeric value for manual setting.

output

logical: Optional output of numerical plot parameters. These can be useful to reproduce similar plots. Default is FALSE.

...

Further plot arguments to pass. xlab must be a vector of length 2, specifying the upper and lower x-axes labels.

Value

returns a plot object and, optionally, a list with plot calculus data.

Function version

0.1.1 (2015-01-25 12:57:49)

Details

The Abanico Plot is a combination of the classic Radial Plot (plot_RadialPlot) and a kernel density estimate plot (e.g. plot_KDE). It allows straightforward visualisation of data precision, error scatter around a user-defined central value and the combined distribution of the values, on the actual scale of the measured data (e.g. seconds, equivalent dose, years). The principle of the plot is shown in Galbraith & Green (1990). The function authors are thankful for the thoughtprovocing figure in this article. The semi circle (z-axis) of the classic Radial Plot is bent to a straight line here, which actually is the basis for combining this polar (radial) part of the plot with any other cartesian visualisation method (KDE, histogram, PDF and so on). Note that the plot allows dispaying two measures of distribution. One is the 2-sigma bar, which illustrates the spread in value errors, and the other is the polygon, which stretches over both parts of the Abanico Plot (polar and cartesian) and illustrates the actual spread in the values themselfes. Since the 2-sigma-bar is a polygon, it can be (and is) filled with shaded lines. To change density (lines per inch, default is 15) and angle (default is 45 degrees) of the shading lines, specify these parameters. See ?polygon() for further help. The Abanico Plot supports other than the weighted mean as measure of centrality. When it is obvious that the data is not (log-)normally distributed, the mean (weighted or not) cannot be a valid measure of centrality and hence central dose. Accordingly, the median and the weighted median can be chosen as well to represent a proper measure of centrality (e.g. centrality = "median.weighted"). Also user-defined numeric values (e.g. from the central age model) can be used if this appears appropriate. The proportion of the polar part and the cartesian part of the Abanico Plot can be modfied for display reasons (plot.ratio = 0.75). By default, the polar part spreads over 75 % and leaves 25 % for the part that shows the KDE graph. The optional parameter layout allows to modify the entire plot more sophisticated. Each element of the plot can be addressed and its properties can be defined. This includes font type, size and decoration, colours and sizes of all plot items. To infer the definition of a specific layout style cf. get_Layout() or type eg. for the layout type "journal" get_Layout("journal"). A layout type can be modified by the user by assigning new values to the list object. It is possible for the z-scale to specify where ticks are to be drawn by using the parameter at, e.g. at = seq(80, 200, 20), cf. function documentation of axis. Specifying tick positions manually overrides a zlim-definition.

References

Galbraith, R. & Green, P., 1990. Estimating the component ages in a finite mixture. International Journal of Radiation Applications and Instrumentation. Part D. Nuclear Tracks and Radiation Measurements, 17 (3), pp. 197-206.

Examples

Run this code

## store original graphics parameters
par.old <- par(no.readonly = TRUE)

## load example data and recalculate to Gray
data(ExampleData.DeValues, envir = environment())
ExampleData.DeValues <- Second2Gray(ExampleData.DeValues, c(0.0438,0.0019))
 
## plot the example data straightforward
plot_AbanicoPlot(data = ExampleData.DeValues)

## now with linear z-scale
plot_AbanicoPlot(data = ExampleData.DeValues,
                 log.z = FALSE)

## now with output of the plot parameters
plot1 <- plot_AbanicoPlot(data = ExampleData.DeValues,
                          output = TRUE)
str(plot1)
plot1$zlim

## now with adjusted z-scale limits
plot_AbanicoPlot(data = ExampleData.DeValues,
                 zlim = c(100, 200))

## now with adjusted x-scale limits
plot_AbanicoPlot(data = ExampleData.DeValues,
                 xlim = c(0, 60))

## now with rug to indicate individual values in KDE part
plot_AbanicoPlot(data = ExampleData.DeValues,
                 rug = TRUE)

## now with a smaller bandwidth for the KDE plot
plot_AbanicoPlot(data = ExampleData.DeValues,
                 bw = 0.01)  

## now with a histogram instead of the KDE plot
plot_AbanicoPlot(data = ExampleData.DeValues,
                 hist = TRUE,
                 kde = FALSE)

## now with a KDE plot and histogram with manual number of bins
plot_AbanicoPlot(data = ExampleData.DeValues,
                 hist = TRUE,
                 breaks = 20)

## now with a KDE plot and a dot plot
plot_AbanicoPlot(data = ExampleData.DeValues,
                 dots = TRUE)

## now with user-defined plot ratio
plot_AbanicoPlot(data = ExampleData.DeValues,
                 plot.ratio = 0.5)

## now with user-defined central value
plot_AbanicoPlot(data = ExampleData.DeValues,
                 central.value = 120)

## now with weighted median as measure of centrality
plot_AbanicoPlot(data = ExampleData.DeValues,
                 centrality = "median.weighted")

## now with median/quartile range as measure of centrality/dispersion
plot_AbanicoPlot(data = ExampleData.DeValues,
                 centrality = "median", 
                 dispersion = "qr")

## now with user-defined green line for minimum age model
CAM <- calc_CentralDose(ExampleData.DeValues,
                        plot = FALSE)

plot_AbanicoPlot(data = ExampleData.DeValues,
                 line = CAM,
                 line.col = "darkgreen",
                 line.label = "CAM")

## now create plot with legend, colour, different points and smaller scale
plot_AbanicoPlot(data = ExampleData.DeValues,
                 legend = "Sample 1",
                 col = "tomato4",
                 bar.col = "peachpuff",
                 pch = "R",
                 cex = 0.8)

## now without 2-sigma bar, polygon, grid lines and central value line
plot_AbanicoPlot(data = ExampleData.DeValues,
                 bar.col = "none",
                 polygon.col = "none",
                 grid.col = "none",
                 y.axis = FALSE,
                 lwd = 0)

## now with direct display of De errors, without 2-sigma bar
plot_AbanicoPlot(data = ExampleData.DeValues,
                 bar.col = "none",
                 ylab = "",
                 y.axis = FALSE,
                 error.bars = TRUE)

## now with user-defined axes labels
plot_AbanicoPlot(data = ExampleData.DeValues,
                 xlab = c("Data error [%]",
                          "Data precision",
                          "Kernel density"),
                 ylab = "Scatter",
                 zlab = "Equivalent dose [Gy]")

## now with minimum, maximum and median value indicated
plot_AbanicoPlot(data = ExampleData.DeValues,
                central.value = 150,
                stats = c("min", "max", "median"))

## now with a brief statistical summary
plot_AbanicoPlot(data = ExampleData.DeValues,
                summary = c("n", "in.ci"))

## now with another statistical summary as subheader
plot_AbanicoPlot(data = ExampleData.DeValues,
                summary = c("mean.weighted", "median"),
                summary.pos = "sub")

## now a plot with two 2-sigma bars for one data set
plot_AbanicoPlot(data = ExampleData.DeValues, 
                 centrality = c(120, 160),
                 mtext = "n = 25")

## now the data set is split into sub-groups, one is manipulated
data.1 <- ExampleData.DeValues[1:15,]
data.2 <- ExampleData.DeValues[16:25,] * 1.3

## now a common dataset is created from the two subgroups
data.3 <- list(data.1, data.2)

## now the two data sets are plotted in one plot
plot_AbanicoPlot(data = data.3)

## now with some graphical modification
plot_AbanicoPlot(data = data.3,
                 col = c("steelblue4", "orange4"),
                 bar.col = c("steelblue3", "orange3"),
                 polygon.col = c("steelblue1", "orange1"),
                 pch = c(2, 6),
                 density = c(10, 20),
                 angle = c(30, 50),
                 summary = c("n", "in.ci"))

## create Abanico plot with predefined layout definition
plot_AbanicoPlot(data = ExampleData.DeValues,
                 layout = "journal")

## now with predefined layout definition and further modifications
plot_AbanicoPlot(data = data.3,
                 layout = "journal",
                 col = c("steelblue4", "orange4"),
                 bar.col = adjustcolor(c("steelblue3", "orange3"), 
                                       alpha.f = 0.5),
                 polygon.col = c("steelblue3", "orange3"))

## for further information on layout definitions see documentation
## of function get_Layout()

## restore original graphical parameters
par(par.old)

Run the code above in your browser using DataLab