raster.modified.ttest
Dutilleul moving window bivariate raster correlation
A bivarate raster correlation using Dutilleul's modified t-test
This function provides a bivariate moving window correlation using the modified t-test to account for spatial autocorrelation. Point based subsampling is provided for computation tractability. The hexagon sampling is recommended as it it good at capturing spatial process that includes nonstationarity and anistropy.
Usage
raster.modified.ttest(
x,
y,
x.idx = 1,
y.idx = 1,
d = "AUTO",
sub.sample = FALSE,
type = "hexagon",
p = 0.1,
size = NULL
)Arguments
- x
x raster for correlation, SpatialPixelsDataFrame or SpatialGridDataFrame object
- y
y raster for correlation, SpatialPixelsDataFrame or SpatialGridDataFrame object
- x.idx
Index for the column in the x raster object
- y.idx
Index for the column in the y raster object
- d
Distance for finding neighbors
- sub.sample
Should a sub-sampling approach be employed (TRUE/FALSE)
- type
If sub.sample = TRUE, what type of sample (random or hexagon)
- p
If sub.sample = TRUE, what proportion of population should be sampled
- size
Fixed sample size
Value
A SpatialPixelsDataFrame or SpatialPointsDataFrame with the following attributes:
corr Correlation
Fstat The F-statistic calculated as [degrees of freedom * unscaled F-statistic]
p.value p-value for the test
moran.x Moran's-I for x
moran.y Moran's-I for y
References
Clifford, P., S. Richardson, D. Hemon (1989), Assessing the significance of the correlationbetween two spatial processes. Biometrics 45:123-134.
Dutilleul, P. (1993), Modifying the t test for assessing the correlation between two spatial processes. Biometrics 49:305-314.
See Also
modified.ttest for test details
Examples
# NOT RUN {
library(gstat)
library(sp)
data(meuse)
data(meuse.grid)
coordinates(meuse) <- ~x + y
coordinates(meuse.grid) <- ~x + y
# GRID-1 log(copper):
v1 <- variogram(log(copper) ~ 1, meuse)
x1 <- fit.variogram(v1, vgm(1, "Sph", 800, 1))
G1 <- krige(zinc ~ 1, meuse, meuse.grid, x1, nmax = 30)
gridded(G1) <- TRUE
G1@data = as.data.frame(G1@data[,-2])
# GRID-2 log(elev):
v2 <- variogram(log(elev) ~ 1, meuse)
x2 <- fit.variogram(v2, vgm(.1, "Sph", 1000, .6))
G2 <- krige(elev ~ 1, meuse, meuse.grid, x2, nmax = 30)
gridded(G2) <- TRUE
G2@data <- as.data.frame(G2@data[,-2])
G2@data[,1] <- G2@data[,1]
corr <- raster.modifed.ttest(G1, G2)
plot(raster::raster(corr,1))
corr.rand <- raster.modifed.ttest(G1, G2, sub.sample = TRUE, type = "random")
corr.hex <- raster.modifed.ttest(G1, G2, sub.sample = TRUE, d = 500, size = 1000)
head(corr.hex@data)
bubble(corr.hex, "corr")
# }
# NOT RUN {
# }