Simulate from the Langevin movement model

This function is based on the Euler approximation.

simLangevinMM(
  beta,
  gamma2 = 1,
  times,
  loc0,
  cov_list = NULL,
  grad_fun = NULL,
  silent = F,
  keep_grad = F,
  b_box = NULL,
  debug = FALSE
)

Arguments

beta	Vector of resource selection coefficients
gamma2	Scalar speed parameter
times	Vector of times of observations. Should have a small time step between each, such that the Euler scheme is valid
loc0	Vector of coordinates (x,y) of initial location
cov_list	List of J (number of covariates) "raster like" elements. A raster like element is a 3 elements list with named elements 1) "x" a vector of increasing x locations (at which the covariate is sampled) 2) "y" a vector of increasing y locations (at which the covariate is sampled) 3) "z" a size(x)*size(y) matrix giving covariate values at location (x, y)
grad_fun	Optional list of functions taking a 2d vector and returning a two 2d vector for the gradient
silent	logical, should simulation advancement be shown?
keep_grad	should gradient values at simulated points be kept? This avoids to compute gradient later on xgrid and ygrid, of dimensions (length(xgrid),length(ygrid),length(beta)).
b_box	a 2x2 matrix with the minimum value on the first line for x and y and the maximum value on the second line
debug	in debug mode, gradient values are kept in grad.csv file stored in the working directory