View source: R/dirichletProcessFit.R
Model options for the Dirichlet Process model. Includes starting values, priors and simulation/MCMC parameters.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 | dirichlet.model.options(iterations = 10000, burnin = 500, thin = 1,
print = 1, start.weights.mixing = NULL, n.clusters = 15,
dropout.estimationTimes = NULL,
dropout.estimationTimes.censored = NULL, dropout.offset = 0,
dp.concentration = NULL, dp.concentration.alpha = NULL,
dp.concentration.beta = NULL, dp.cluster.sigma = NULL,
dp.cluster.sigma.nu0 = NULL, dp.cluster.sigma.T0 = NULL,
dp.dist.mu0 = NULL, dp.dist.mu0.mb = NULL, dp.dist.mu0.Sb = NULL,
dp.dist.sigma0 = NULL, dp.dist.sigma0.nub = NULL,
dp.dist.sigma0.Tb = NULL, betas.covariates = NULL,
betas.covariates.mu = NULL, betas.covariates.sigma = NULL,
sigma.error = NULL, sigma.error.tau = NULL,
density.intercept.domain = NULL, density.slope.domain = NULL,
censoring.var = NULL)
|
iterations |
number of iterations for the MCMC simulation |
burnin |
burn in period for the simulation, i.e. the number of iterations to throw away at the beginning of the simulation |
thin |
thinning interval, i.e. if thin=n, only keep every nth iteration |
print |
printing interval, i.e. if print=n, print a counter on every nth iteration |
dp.concentration |
prior value for the concentration parameter of the Dirichlet process |
dp.concentration.alpha |
Shape parameter for the hyperprior Gamma distribution of the concentration parameter of the Dirichlet process |
dp.concentration.beta |
Rate parameter for the hyperprior Gamma distribution of the concentration parameter of the Dirichlet process |
dp.cluster.sigma |
prior for the common cluster covariance |
dp.cluster.sigma.nu0 |
Degrees of freedom for the hyperprior inverse Wishart distribution of the common cluster covariance |
dp.cluster.sigma.T0 |
Scale matrix for the hyperprior inverse Wishart distribution of the common cluster covariance |
dp.dist.mu0 |
prior mean for the baseline distribution of the Dirichlet process |
dp.dist.mu0.mb |
Mean for the hyperprior Gaussian distribution of the mean of the baseline distribution of the Dirichlet process |
dp.dist.mu0.Sb |
Covariance for the hyperprior Gaussian distribution of the mean of the baseline distribution of the Dirichlet process |
dp.dist.sigma0 |
prior covariance for the baseline distribution of the Dirichlet process |
dp.dist.sigma0.nub |
Degrees of freedom for the hyperprior inverse Wishart distribution of the covariance of the baseline distribution of the Dirichlet process |
dp.dist.sigma0.Tb |
Scale matrix for the hyperprior inverse Wishart distribution of the covariance of the baseline distribution of the Dirichlet process |
betas.covariates |
prior value for covariate regression coefficients |
betas.covariates.mu |
Prior mean for the covariate regression coefficients |
betas.covariates.sigma |
Prior covariance for the covariate regression coefficients |
sigma.error |
Prior for the residual error (Gaussian outcomes only) |
sigma.error.tau |
Hyperprior for the residual error (Gaussian outcomes only) |
density.intercept.domain |
vector of values at which to calculate the density of the random intercepts. If NULL, no density will be calculated |
density.slope.domain |
vector of values at which to calculate the density of the random slopes. If NULL, no density will be calculated |
censoring.var |
column of data set containing an indicator for administrative censoring (optional). 1=dropout time censored, 0=actual dropout time observed. |
numClusters |
number of clusters for the Dirichlet Process stick breaking model |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.