dynfrail_dist: Distribution parameters for dynfrail
In tbalan/dynfrail: Fitting Dynamic Frailty Models with the EM Algorithm

Description Usage Arguments Details Value See Also Examples

View source: R/dynfrail_arguments.R

Distribution parameters for dynfrail

1 2	dynfrail_dist(dist = "gamma", theta = 2, pvfm = -1/2, lambda = 0.1, n_ints = NULL, times = NULL)

`dist`	One of 'gamma', 'stable' or 'pvf'.
`theta`	Frailty distribution parameter. Must be >0.
`pvfm`	Only relevant if `dist = 'pvf'` is used. It determines which PVF distribution should be used. Must be larger than -1 and not equal to 0.
`lambda`	Frailty autocorrelation parameter. Must be >0.
`n_ints`	For piece-wise constant frailty, the number of intervals. With `n_ints = 0`, the classical shared frailty scenario is obtained.
`times`	A vector of time points which determine the piecewise-constant interval for the frailty. Overrides `n_ints`.

The theta and lambda arguments must be positive. In the case of gamma or PVF, theta is the inverse of the frailty variance, i.e. the larger the theta is, the closer the model is to a Cox model. When dist = "pvf" and pvfm = -0.5, the inverse Gaussian distribution is obtained. For the positive stable distribution, the γ parameter of the Laplace transform is θ / (1 + θ), with the alpha parameter fixed to 1.

An object of the type dynfrail_dist, which is mostly used to denote the supported frailty distributions in a consistent way.

dynfrail, dynfrail_control

dynfrail_dist()
# Compound Poisson distribution:
dynfrail_dist(dist = 'pvf', theta = 1.5, pvfm = 0.5)
# Inverse Gaussian distribution:
dynfrail_dist(dist = 'pvf')