Pimage: Bin MCMC samples

In KateGoodenough/RoL-SGAT: Solar/Satellite Geolocation for Animal Tracking

Description

Create object to store binned images.

Usage

Pimage(x, ...)

## S3 method for class 'POSIXct'
Pimage(x, type = c("primary", "intermediate"),
  pimg = NULL, grid = NULL, proj = NULL, ...)

## Default S3 method:
Pimage(x, type = c("primary", "intermediate"),
  pimg = NULL, grid = NULL, proj = NULL, ...)

Arguments

x	vector of POSIXct date-times for locations
...	arguments passed to chain.bin
type	character, samples to bin, "primary" or "intermediate"
pimg	a Pimage object to accept these samples, if not supplied one will be created base on the inputs
grid	object to use as a template for grid specification, can be a anything accepted by raster or a Pimage object (any previous binning will be reset), see Details
proj	a character string of projection arguments, see Details

Details

These functions provide a data structure tools to store binned samples generated by the Metropolis samplers.

These functions provide spatial binning of samples. A spatial summary image is stored separately for each time step and may be mosaiced into the entire study region. Separate summaries may be combined to create a multiple-track summary.

If pimg is supplied grid and proj are ignored and binning is added to the existing pimg. If pimg or is not supplied grid is used to build one with the details from the fit object, and proj is ignored. If only proj is supplied a grid is build using that projection and the details from the fit object.

The proj argument should be a PROJ.4 string, see projection and CRS, or an incomplete PROJ.4 name string. If the string consists only of the projection family name then a central coordinate is calculated from the samples.See rgdal::projInfo("proj")$name for candidate strings, and http://www.spatialreference.org for more details.

Value

Pimage

Examples

## Not run: 
## Brownian motion tethered at each end
brownian.bridge <- function(n, r) {
  x <- cumsum(rnorm(n, 0, 1))
  x <- x - (x[1] + seq(0, 1, length=n) * (x[n] - x[1]))
  r * x
}

## Number of days and number of obs
days <- 50
n <- 200

x <- rgamma(n, 3)
x <- cumsum(x)
x <- x/x[n]

b.scale <- 0.6
r.scale <- sample(c(0.1, 2, 10.2), n, replace=TRUE,
                  prob=c(0.8, 0.18, 0.02))
set.seed(71)

tms <- ISOdate(2001, 1, 1) + trunc(days * 24 * 60 * 60 *x)
lon <- 120 + 20 * sin(2 * pi * x) +
brownian.bridge(n, b.scale) + rnorm(n, 0, r.scale)
lat <- -40 + 10 *(sin(3 * 2 * pi * x) + cos(2 * pi * x) - 1) +
brownian.bridge(n, b.scale) + rnorm(n, 0, r.scale)
x0 <- cbind(lon, lat)
z0 <- trackMidpts(x0)
n3 <- 1500
fx <- list(x = list(array(NA_real_, c(length(lon), 2L, n3))),
           z = list(array(NA_real_, c(length(lon)-1L, 2L, n3))),
           model = list(time = tms))
fx$x[[1L]][,,1L] <- x0
fx$z[[1L]][,,1L] <- z0
for (i in seq(n3)[-1L]) {
   fx$x[[1L]][,,i] <- jitter(x0, factor = 8L)
   fx$z[[1L]][,,i] <- jitter(z0, factor = 12L)
}

g <- raster(extent(x0) + 5, nrows = 350, ncols = 375, crs = "+proj=longlat +datum=WGS84")
px <- Pimage(fx, grid = g)
pz <- Pimage(fx, type = "intermediate", grid = g)

for (i in seq(n3)[-1L]) {
fx$x[[1L]][,,i] <- jitter(x0, factor = 8L)
fx$z[[1L]][,,i] <- jitter(z0, factor = 12L)
}

px2 <- Pimage(fx, pimg = px)
pz2 <- Pimage(fx, type = "intermediate", pimg = pz)

px3 <- Pimage(fx, grid = g)
pz3 <- Pimage(fx, grid = g, type  = "intermediate")

## first
px$p[[80]]
## this should be the sum of first and last
px2$p[[80]]
## last
px3$p[[80]]

## End(Not run)

KateGoodenough/RoL-SGAT documentation built on June 11, 2019, 1:29 p.m.