Title: | Conditional Aalen-Johansen Estimation |
---|---|
Description: | Provides the conditional Nelson-Aalen and Aalen-Johansen estimators. The methods are based on Bladt & Furrer (2023), in preparation. |
Authors: | Martin Bladt [aut, cre], Christian Furrer [aut] |
Maintainer: | Martin Bladt <[email protected]> |
License: | GPL (>= 2) |
Version: | 1.0 |
Built: | 2024-12-19 06:40:13 UTC |
Source: | CRAN |
Compute the conditional Aalen-Johansen estimator.
aalen_johansen( data, x = NULL, a = NULL, p = NULL, alpha = 0.05, collapse = FALSE )
aalen_johansen( data, x = NULL, a = NULL, p = NULL, alpha = 0.05, collapse = FALSE )
data |
A list of trajectory data for each individual. |
x |
A numeric value for conditioning. |
a |
A bandwidth. Default uses an asymmetric version using alpha. |
p |
An integer representing the number of states. The absorbing state is last. |
alpha |
A probability around the point x, for asymmetric sub-sampling. |
collapse |
Logical, whether to collapse the last state of the model. |
A list containing the Aalen-Johansen estimator, the Nelson-Aalen estimator, and related quantities.
Calculate the product integral of a matrix function
prodint(start, end, step_size, lambda)
prodint(start, end, step_size, lambda)
start |
Start time. |
end |
End time. |
step_size |
Step size of the grid. |
lambda |
A given matrix function. |
The product integral of the given matrix function.
Simulate the path of a time-inhomogeneous (semi-)Markov process until a maximal time
sim_path(i, rates, dists, t = 0, u = 0, tn = Inf, abs = numeric(0), bs = NA)
sim_path(i, rates, dists, t = 0, u = 0, tn = Inf, abs = numeric(0), bs = NA)
i |
The initial state, integer. |
rates |
The total transition rates out of states, a function with arguments state (integer), time (numeric), and duration (numeric) returning a rate (numeric). |
dists |
The distribution of marks, a function with arguments state (integer), time (numeric), and duration (numeric) returning a probability vector. |
t |
The initial time, numeric. |
u |
The initial duration (since the last transition), numeric. By default equal to zero |
tn |
The maximal time, numeric. By default equal to inifinity |
abs |
Vector indicating which states are absorbing. By default the last state is absorbing. |
bs |
Vector of upper bounds on the total transition rates. By default the bounds are determined using optimize, which might only identify a local maximum. |
A list concerning jump times and states, with the first time being the initial time t and state and the last time being tn (if not absorbed)
jump_rate <- function(i, t, u){if(i == 1){3*t} else if(i == 2){5*t} else{0}} mark_dist <- function(i, s, v){if(i == 1){c(0, 1/3, 2/3)} else if(i == 2){c(1/5, 0, 4/5)} else{0}} sim <- sim_path(sample(1:2, 1), t = 0, tn = 2, rates = jump_rate, dists = mark_dist) sim
jump_rate <- function(i, t, u){if(i == 1){3*t} else if(i == 2){5*t} else{0}} mark_dist <- function(i, s, v){if(i == 1){c(0, 1/3, 2/3)} else if(i == 2){c(1/5, 0, 4/5)} else{0}} sim <- sim_path(sample(1:2, 1), t = 0, tn = 2, rates = jump_rate, dists = mark_dist) sim