Interpolate clusters from STRUCTURE, TESS or MATRIX runs spatially using coordinates.
plotRunsInterpolate(datafile = NULL, coordsfile = NULL, method = "krig",
duplicate = "mean", idwpower = 2, clusters = NA, gridsize = 60,
imgoutput = "join", colours = NA, nrow = NA, ncol = NA,
exportplot = TRUE, imgtype = "png", height = NA, width = NA,
units = "cm", res = 200, showaxis = FALSE, addpoints = TRUE,
pointcol = "grey10", pointtype = "+", pointsize = 4, legend = TRUE,
legendpos = c(1, 1), legendjust = c(1, 1), legendsize = NA,
legendtextsize = NA, dataout = FALSE)One STRUCTURE, TESS or MATRIX output file. Input is either a character
or a dataframe. If character, then a path pointing to location of the datafile. Can use
choose.files(). If a dataframe, then an output from runsToDfStructure(),
runsToDfTess() or runsToDfMatrix().
A character or a dataframe. If character, then a path pointing to location of the coordsfile. It must be a tab-delimited text file with x and y coordinates of the samples. The number of rows must be equal to the number of samples in datafile. The coordsfiles must have no header and 2 columns in the order: x (latitude) and then y (longitude). Coordinates must be in standard longitude latitude (LL) decimals.
A character indicating the method employed for interpolation. Options are "bilinear", "bicubic", "krig" (Kriging), "idw" (Inverse distance weighting) or "nn" (nearest neighbour). See Details for more information.
A character indicating how to deal with duplicate spatial locations. This is only applicable to 'bilinear' and 'bicubic' methods. Options are "error" (error message if duplicate points), "strip" (remove all duplicate points), "mean" (mean of duplicate points), "median" (median of duplicate points).
A numeric indicating the power of inverse distance weighting if method is set to "idw".
A numeric or numeric vector indicating the clusters to plot. If NA, all clusters are plotted. For ex. If set to 2, cluster 2 is plotted. If set to 2:4, clusters 2, 3 and 4 are plotted. If set to c(1,4,5), clusters 1, 4 and 5 are plotted.
A numeric indicating the size of the image grid on which interpolation is to be carried out. Set to 60 by default. Higher values produces less pixelated grids, but more computationally intensive.
A character. To plot each cluster as a seperate figure, set to "sep". To plot multiple clusters in a single figure, set to "join". If number of clusters is less than 2, then automatically set to "sep".
A vector of colours. R colour names or hexadecimal values. Set to 9 value 'Blues' palette from RColorBrewer by default.
A numeric indicating the number of rows of plots in a joined plot. Determined automatically
if number of plots <20 and nrow = NA.
A numeric indicating the number of columns of plots in a joined plot. Determined automatically
if number of plots <20 and ncol = NA.
A logical. If set to FALSE, no image is exported.
A character indicating the export format for figures. Options are "png", "jpeg" or "pdf". If pdf, height and width must be in inches and res argument is ignored (set to 300).
A numeric indicating the height of export figure in cm unless units are changed.
A numeric indicating the width of export figure in cm unless units are changed.
A character indicating units of measurement for figure dimensions. Set to 'cm' by default.
A numeric indicating the pixel resolution of the export image. Set to 200 by default.
A logical. If TRUE, then axis text, axis ticks and plot border are plotted.
A logical. If TRUE, then sample coordinates are overplotted on interpolated grid.
Colour of sample points. An R colour or hexadecimal value.
The shape/pch of sample points. A numeric or a character.
The size of sample points. A number usually 0.4,0.8,1,3 etc.
A logical. If TRUE, the legend for the colours is plotted.
A character of 2-value numeric vector indicating the position of the legend. If "right","left","top" or "bottom", then, legend is plotted outside the plot area. To plot inside plot area use a 2 vale vector. If a vector like c(1,1), first value denotes x-axis from 0 to 1 and second value denotes y-axis from 0 to 1. For ex. to plot in bottom left corner, use c(0,0).
The x and y axis justification of the legend. A 2 value vector.
A numeric indicating the size of the legend in cm. Usually values like 0.5,0.7,1.2 etc. The legendsize does not control the text in the legend.
A numeric indicating the size of the text in the legend.
A logical. If set to TRUE, a list of one or more ggplot gtable elements are returned.
This output can be modified using ggplot themes() for more figure control if required.
If dataout = TRUE, a list of one or more ggplot gtable output is returned for more theme
control if required.
The "bilinear", "bicubic", "idw" and "nn" are essentially direct interpolation
between spatial points. The "krig" option is predictive rather than direct interpolation.
The kriging function is same function provided by the TESS authors in their R script.
The function uses great circle distances rdist.earth() from fields package to
determine theta. Therefore coordinates must be in LL and not UTM.
For more details of methods, see R package akima function interp for "bilinear"
and "bicubic" methods, see R package fields function Krig for "krig" method, see
R package spatstat function idw for "idw" and function nnmark for "nn" method.
The model for "krig" is automatically determined and may produce warning messages if
the GCV algorithm does not converge optimally. This shouldn't be an issue for exploratory
analyses. All methods require full coordinate data. No missing data allowed in coordsfile.
# NOT RUN {
#tess files
tlist <- list.files(path=system.file("files/tess",package="pophelper"),full.names=TRUE)
cd1 <- system.file("files/coords75.txt",package="pophelper")
plotRunsInterpolate(datafile=tlist[2],coordsfile=cd1)
# }
Run the code above in your browser using DataLab