vonMisesMixtureCreate: Create a von Mises mixing distribution
In keesmulder/circbayes: Bayesian Analysis of Circular Data

Create a von Mises mixing distribution

1 2	vonMisesMixtureCreate(priorParameters, priorMeanMethod = "integrate", n_samp = 3)

`priorParameters`	Prior parameters for the base measure which are, in order, (mu_0, R_0, c). The prior mean can be set to `NA` to be unknown. It is then dealt with using a method selected by `priorMeanMethod`.
`priorMeanMethod`	Character; Method to deal with with the prior mean mu_0, if it is given not given but set to `NA`. If `"datadependent"`, the prior is assumed to be uninformative for the mean direction, but the prior resultant length R_0 is used 'as is', that is the posterior resultant length is R_n = R + R_0. This implictly assumes that the prior mean is the data mean direction \barθ, which is why it is called 'data dependent'. If `"integrate"`, the posterior parameters are computed averaged over a uniform distribution on the prior mean. If `"sample"`, a prior mean is sampled uniformly at each iteration.
`n_samp`	Number of samples to generate from the Gibbs sampler before accepting the sample as an i.i.d. sample from the posterior. Higher values mean we can be more certain that the sampler works correctly, at the cost of computational time. The sampler mixes very fast, but different values can be attempted by means of sensitivity. As a general recommendation, 2 is only approximately correct, 3 is close, and 4 performs almost as good as 100 or more.