calRaster: function to rescale isoscape

In SPATIAL-Lab/isorig: Rescale isoscape using known-origin tissue isotope values, and assign origin of unknown samples with isotopic data

Description

This function relies on known-origin tissue data to calibrate between isotope values in precipitation and those in the tissue using simple regression to rescale the precipitation raster into tissue values and generate mean and standard deviation rasters. This will output plots as pdf including regression model and rescaled isoscape mean and standard deviation in output folder under working directory.

Usage

calRaster(known, isoscape, mask = NULL, sdMethod = 2, interpMethod = 2, NA.value = NA, ignore.NA = T, genplot = T, savePDF =  T, verboseLM = T)

Arguments

known	SpatialPointsDataFrame. known-origin data that should contain only one feature: tissue isotope value. The coordinate reference system must be the same as isoscape and mask. Optional known-origin data can be queried using the knownOrig function.
isoscape	raster. isoscape raster with two layers. The first one is the mean and the second one is either standard deviation or 2 standard deviation. Its coordinate reference system must be the same as input of known and mask.
mask	SpatialPolygonsDataFrame. This mask will constrain the isoscape area. If this is not provided, default of mask of isoscape in this argument is used. Its coordinate reference system must be the same as input of known and isoscape.
sdMethod	Uncertainty in the isoscape is standard deviation (1) or two standard deviation (2, default)
interpMethod	numeric. 1 or 2. Designate oen of two methods for extracting values from isoscape raster based on known-origin position. If 1, values for the cell in which a point falls are returned. If 2 (default) the returned values are interpolated (bilinear interpolation) from the values of the four nearest raster cells.
NA.value	NA [default] or numeric. Identify the value that was used to represent an absence of data in the isoscape raster map. Commonly used values include NA, -9999, 9999, etc. If you are not sure what value was used in your isoscape, you can load the isoscape raster in R and run click(raster) to click the place without value.
ignore.NA	logical. (TRUE or FALSE [default]). If NA values are extracted from isoscape at the location of known-origin, do you want to ignore these values and proceed with calRaster function?
genplot	logical. Plot the results (TRUE [default]) or not (FALSE)).
savePDF	logical. Save the results as pdf (TRUE [default]) or not (FALSE)).
verboseLM	logical. Print out the linear regression results (TRUE [default]) or not (FALSE)).

Value

isoscape.rescale	Rescaled isoscape based on known origin data and isosape, two layer raster, one is mean the other is standard deviation which is a universal value for the map from error of linear regression model.
lm.data	known origin data and correlated isoscape value used for linear regression model
lm.model	linear regression model

See Also

pdRaster

Examples

data("naMap")
data("d2h_world")
# the columns of d should be longitude, latitude, and tissue isotope
d <- subOrigData(taxon = c("Homo sapiens"), mask = naMap)
r <- calRaster(known = d, isoscape = d2h_world, mask = naMap)

SPATIAL-Lab/isorig documentation built on Aug. 13, 2019, 11:02 p.m.