Minor functions.
getMeanSD(xy)
maskarea(mask, sessnum = 1)
masklength(mask, sessnum = 1)
masksize(mask, sessnum = 1)
edist(xy1, xy2)
nedist(xy1, xy2, mask, inf = Inf, ...)
rlnormCV(n, mean, cv)
For getMeanSD, a dataframe with columns `x' and `y' and two
rows, mean and SD.
For maskarea, the summed area of mask cells in hectares (ha).
For masklength, the summed length of mask cells in kilometers (km).
For masksize, whichever of area or length is appropriate.
For edist and nedist, a matrix with dim = c(nrow(xy1), nrow(xy2)).
For rlnormCV a vector of random deviates.
2-column matrix or dataframe
2-column matrix or dataframe
2-column matrix or dataframe
mask or linearmask object
integer; for multi-session masks, the number of the session
numeric value to use for +infinity
other arguments for transition
number of observations
mean on natural scale
coefficient of variation on natural scale
getmeanSD is used by make.mask to standardize
mask coordinates.
For masklength the input should be a linear mask from secrlinear.
edist computes the Euclidean distance between each point in xy1
and each point in xy2. (This duplicates the functionality of `rdist'
in package fields).
nedist computes the non-Euclidean distance between each point
in xy1 and each point in xy2, in two dimensions. The calculation uses
gdistance (van Etten 2017; see also Csardi & Nepusz 2006): a
transition layer is formed representing the connections between
adjacent points in mask. By default, points within a 16-point
neighbourhood are considered `adjacent'. Distances are obtained by
Dijkstra's (1959) algorithm as least cost paths through the graph of
all points in the mask.
nedist has some subtle options. If `mask' is missing then the
transition layer will be formed from `xy2'. If `mask' has a covariate
named `noneuc' then this will be used to weight distances. The ...
argument of nedist allows the user to vary arguments of
transition (defaults transitionFunction =
mean and directions = 16). Be warned this can lead to unexpected
results! Point pairs that are completely separated receive the
distance +Inf unless a finite value is provided for the argument
`inf'. See
secr-noneuclidean.pdf
for uses of nedist.
rlnormCV is a wrapper for rlnorm that computes its
meanlog and sdlog arguments from the mean and CV on the natural scale:
\(\mbox{sdlog} = \sqrt{\log(\mbox{CV}^2 + 1)}\), and
\(\mbox{meanlog} = \log(\mbox{mean}) - \mbox{sdlog}^2/2\).
Dijkstra, E. W. (1959) A note on two problems in connexion with graphs. Numerische Mathematik, 1, 269--271.
Csardi, G. and Nepusz, T. (2006) The igraph software package for complex network research. InterJournal, 1695. https://igraph.org
van Etten, J. (2017) R package gdistance: Distances and routes on geographical grids. Journal of Statistical Software, 76(1), 1--21. tools:::Rd_expr_doi("10.18637/jss.v076.i13")