likinvbat: Likelihood of the inverse Batschelet distribution
In keesmulder/flexcircmix: Fit Mixtures of Flexible Circular Distributions

Description Usage Arguments Value Functions Examples

Two functions to obtain the inverse Batschelet likelihood of a set parameters, given a data set of angles.

1
2
3

likinvbat(x, mu, kp, lam, weights = rep(1, length(x)), log = TRUE)

likfuninvbat(x, weights = rep(1, length(x)), log = TRUE)

`x`	A set of angles in radians.
`mu`	A mean direction, in radians.
`kp`	Numeric, > 0,the concentration parameter.
`lam`	The shape parameter (peakedness), -1 < `lam` < 1.
`weights`	A vector of length `length(x)`, which gives importance weights to be used for x.
`log`	If `TRUE` (the default), the log-likelihood is used.

likinvbat returns a value, the likelihood given the data and parameters. likfuninvbat returns a function of mu, kp and lam, which can be evaluated later for a given set of parameters.

likfuninvbat: Return a likelihood function.

vm_data <- circglmbayes::rvmc(10, 1, 5)
llfun <- likfuninvbat(vm_data)

# log-likelihood value of true parameters.
llfun(mu = 1, kp = 5, lam = 0)

# Plot the conditional log-likelihood.
kp_conditional_ll <- Vectorize(function(x) llfun(mu = 1, kp = x, lam = 0))
curve(kp_conditional_ll, 0, 20)