NegativeLogLikelihood: Computes negative log-likelihood of moveDs model
In r-glennie/moveds: Fit Distance Sampling Models with Animal Movement

Computes negative log-likelihood of moveDs model

NegativeLogLikelihood(working_parameter, start, data, transdat,
  auxiliary_data, delta, num_cells, T, ymax, buffer, movement_data,
  fixed_sd = 0, hzfn = 1L, move_method = 1L, print = FALSE,
  con = 100)

`working_parameter`	unconstrained version of parameter vector containing (detection shape, detection scale, diffusion sd)
`start`	start value for parameters on natural scale
`data`	matrix with (trans id, grid cell,t) distance sampling survey data (assumed to be ordered by transect and time)
`transdat`	matrix with (stripsize(1), numcells in y, totaltimestep, number of observations)
`auxiliary_data`	vector containing (area x extent, area y extent, strip width, transect_type)
`delta`	vector of (dx, dt) spacetime increments
`num_cells`	number of cells in (total space, x-direction, y-direction)
`T`	total time of survey for longest transect
`ymax`	maximum length of a transect
`buffer`	buffer distance
`movement_data`	field object where each component represents an individual path and contains a matrix where each row is an observed location (x,y,t)
`fixed_sd`	if move_method = 2
`hzfn`	hazard function code (see ?hazardfns)
`move_method`	0 = 2d CDS model, 1 = 2d MDS model (movement estimated), 2 = 2d MDS model (movement fixed)
`print`	if TRUE then print likelihood and parmeters after evaluation
`con`	parameters are constrained to be between 1/con * start value and con * start value