Package 'psbcGroup' reference manual

Title:	Penalized Parametric and Semiparametric Bayesian Survival Models with Shrinkage and Grouping Priors
Description:	Algorithms to implement various Bayesian penalized survival regression models including: semiparametric proportional hazards models with lasso priors (Lee et al., Int J Biostat, 2011 <doi:10.2202/1557-4679.1301>) and three other shrinkage and group priors (Lee et al., Stat Anal Data Min, 2015 <doi:10.1002/sam.11266>); parametric accelerated failure time models with group/ordinary lasso prior (Lee et al. Comput Stat Data Anal, 2017 <doi:10.1016/j.csda.2017.02.014>).
Authors:	Kyu Ha Lee, Sounak Chakraborty, Harrison Reeder, (Tony) Jianguo Sun
Maintainer:	Kyu Ha Lee <[email protected]>
License:	GPL (>= 2)
Version:	1.7
Built:	2024-11-05 06:15:23 UTC
Source:	CRAN

Function to Fit the Penalized Parametric Bayesian Accelerated Failure Time Model with Group Lasso Prior

Description

Penalized parametric Bayesian accelerated failure time model with group lasso prior is implemented to analyze survival data with high-dimensional covariates.

Usage

aftGL(Y, data, grpInx, hyperParams, startValues, mcmc)
aftGL(Y, data, grpInx, hyperParams, startValues, mcmc)

Arguments

`Y`	a data.frame containing univariate time-to-event outcomes from `n` subjects. It is of dimension $n\times 2$ : the columns correspond to $y$ , $\delta$ .
`data`	a data.frame containing $p$ covariate vectors from `n` subjects. It is of dimension $n\times p$ .
`grpInx`	a vector of $p$ group indicator for each variable
`hyperParams`	a list containing hyperparameter values in hierarchical models: (`nu0`, `sigSq0`): hyperparameters for the prior of $\sigma^2$ ; (`alpha0`, `h0`): hyperparameters for the prior of $\alpha$ ; (`rLam`, `deltaLam`): hyperparameters for the prior of $\lambda^2$ .
`startValues`	a list containing starting values for model parameters. See Examples below.
`mcmc`	a list containing variables required for MCMC sampling. Components include, `numReps`, total number of scans; `thin`, extent of thinning; `burninPerc`, the proportion of burn-in. See Examples below.

Value

aftGL returns an object of class aftGL.

Author(s)

Kyu Ha Lee, Sounak Chakraborty, (Tony) Jianguo Sun

References

Lee, K. H., Chakraborty, S., and Sun, J. (2017). Variable Selection for High-Dimensional Genomic Data with Censored Outcomes Using Group Lasso Prior. Computational Statistics and Data Analysis, Volume 112, pages 1-13.

Examples


# generate some survival data	
	set.seed(204542)
	
	p = 20
	n = 200
	logHR.true <- c(rep(4, 10), rep(0, (p-10)))	

	CovX<-matrix(0,p,p)

	for(i in 1:10){
		for(j in 1:10){
			CovX[i,j] <- 0.3^abs(i-j)
			}
		}
		
	diag(CovX) <- 1
	
	data	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)	
	pred <- as.vector(exp(rowSums(scale(data, center = FALSE, scale = 1/logHR.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	tcen 		<- pmin(t, cen)
	di 		<- as.numeric(t <= cen)
	
	n <- dim(data)[1]
	p <- dim(data)[2]

	Y <- data.frame(cbind(tcen, di))
	colnames(Y) <- c("time", "event")

	grpInx <- 1:p
	K <- length(unique(grpInx))
	
	############################
	hyperParams <- list(nu0=3, sigSq0=1, alpha0=0, h0=10^6, rLam=0.5, deltaLam=2)

	############################
	startValues <- list(alpha=0.1, beta=rep(1,p), sigSq=1, tauSq=rep(0.4,p), lambdaSq=5,
	 				w=log(tcen))

	############################	
	mcmc <- list(numReps=100, thin=1, burninPerc=0.5)
	
	############################
	fit <- aftGL(Y, data, grpInx, hyperParams, startValues, mcmc)
## Not run:   
vs <- VS(fit, X=data)

## End(Not run)
# generate some survival data	
	set.seed(204542)
	
	p = 20
	n = 200
	logHR.true <- c(rep(4, 10), rep(0, (p-10)))	

	CovX<-matrix(0,p,p)

	for(i in 1:10){
		for(j in 1:10){
			CovX[i,j] <- 0.3^abs(i-j)
			}
		}
		
	diag(CovX) <- 1
	
	data	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)	
	pred <- as.vector(exp(rowSums(scale(data, center = FALSE, scale = 1/logHR.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	tcen 		<- pmin(t, cen)
	di 		<- as.numeric(t <= cen)
	
	n <- dim(data)[1]
	p <- dim(data)[2]

	Y <- data.frame(cbind(tcen, di))
	colnames(Y) <- c("time", "event")

	grpInx <- 1:p
	K <- length(unique(grpInx))
	
	############################
	hyperParams <- list(nu0=3, sigSq0=1, alpha0=0, h0=10^6, rLam=0.5, deltaLam=2)

	############################
	startValues <- list(alpha=0.1, beta=rep(1,p), sigSq=1, tauSq=rep(0.4,p), lambdaSq=5,
	 				w=log(tcen))

	############################	
	mcmc <- list(numReps=100, thin=1, burninPerc=0.5)
	
	############################
	fit <- aftGL(Y, data, grpInx, hyperParams, startValues, mcmc)
## Not run:   
vs <- VS(fit, X=data)

## End(Not run)

Function to Fit the Penalized Parametric Bayesian Accelerated Failure Time Model with Group Lasso Prior for Left-Truncated and Interval-Censored Data

Description

Penalized parametric Bayesian accelerated failure time model with group lasso prior is implemented to analyze left-truncated and interval-censored survival data with high-dimensional covariates.

Usage

aftGL_LT(Y, X, XC, grpInx, hyperParams, startValues, mcmcParams)
aftGL_LT(Y, X, XC, grpInx, hyperParams, startValues, mcmcParams)

Arguments

`Y`	Outcome matrix with three column vectors corresponding to lower and upper bounds of interval-censored data and left-truncation time
`X`	Covariate matrix $p$ covariate vectors from `n` subjects. It is of dimension $n\times p$ .
`XC`	Matrix for confound variables: $q$ variable vectors from `n` subjects. It is of dimension $n\times q$ .
`grpInx`	a vector of $p$ group indicator for each variable
`hyperParams`	a list containing hyperparameter values in hierarchical models: (`a.sigSq`, `a.sigSq`): hyperparameters for the prior of $\sigma^2$ ; (`mu0`, `h0`): hyperparameters for the prior of $\mu$ ; (`v`): hyperparameter for the prior of $\beta_C$ .
`startValues`	a list containing starting values for model parameters. See Examples below.
`mcmcParams`	a list containing variables required for MCMC sampling. Components include, `numReps`, total number of scans; `thin`, extent of thinning; `burninPerc`, the proportion of burn-in. See Examples below.

Value

aftGL_LT returns an object of class aftGL_LT.

Author(s)

Kyu Ha Lee, Harrison Reeder

References

Reeder, H., Haneuse, S., Lee, K. H. (2024+). Group Lasso Priors for Bayesian Accelerated Failure Time Models with Left-Truncated and Interval-Censored Data. under review

Examples


## Not run: 

data(survData)
X <- survData[,c(4:5)]
XC <- NULL

n <- dim(survData)[1]
p <- dim(X)[2]
q <- 0

c0 <- rep(0, n)
yL <- yU <- survData[,1]
yU[which(survData[,2] == 0)] <- Inf
Y <- cbind(yL, yU, c0)

grpInx <- 1:p
K <- length(unique(grpInx))

#####################
## Hyperparameters

a.sigSq= 0.7
b.sigSq= 0.7

mu0 <- 0
h0 <- 10^6

v = 10^6

hyperParams <- list(a.sigSq=a.sigSq, b.sigSq=b.sigSq, mu0=mu0, h0=h0, v=v)

###################
## MCMC SETTINGS

## Setting for the overall run
##
numReps    <- 100
thin    <- 1
burninPerc <- 0.5

## Tuning parameters for specific updates
##

L.beC <- 50
M.beC <- 1
eps.beC <- 0.001

L.be <- 100
M.be <- 1
eps.be <- 0.001

mu.prop.var    <- 0.5
sigSq.prop.var    <- 0.01

##

mcmcParams <- list(run=list(numReps=numReps, thin=thin, burninPerc=burninPerc),
tuning=list(mu.prop.var=mu.prop.var, sigSq.prop.var=sigSq.prop.var,
L.beC=L.beC, M.beC=M.beC, eps.beC=eps.beC,
L.be=L.be, M.be=M.be, eps.be=eps.be))

#####################
## Starting Values

w        <- log(Y[,1])
mu     <- 0.1
beta     <- rep(2, p)
sigSq    <- 0.5
tauSq <- rep(0.4, p)
lambdaSq <- 100
betaC     <- rep(0.11, q)

startValues <- list(w=w, beta=beta, tauSq=tauSq, mu=mu, sigSq=sigSq,
lambdaSq=lambdaSq, betaC=betaC)

fit <- aftGL_LT(Y, X, XC, grpInx, hyperParams, startValues, mcmcParams)


## End(Not run)
## Not run: 

data(survData)
X <- survData[,c(4:5)]
XC <- NULL

n <- dim(survData)[1]
p <- dim(X)[2]
q <- 0

c0 <- rep(0, n)
yL <- yU <- survData[,1]
yU[which(survData[,2] == 0)] <- Inf
Y <- cbind(yL, yU, c0)

grpInx <- 1:p
K <- length(unique(grpInx))

#####################
## Hyperparameters

a.sigSq= 0.7
b.sigSq= 0.7

mu0 <- 0
h0 <- 10^6

v = 10^6

hyperParams <- list(a.sigSq=a.sigSq, b.sigSq=b.sigSq, mu0=mu0, h0=h0, v=v)

###################
## MCMC SETTINGS

## Setting for the overall run
##
numReps    <- 100
thin    <- 1
burninPerc <- 0.5

## Tuning parameters for specific updates
##

L.beC <- 50
M.beC <- 1
eps.beC <- 0.001

L.be <- 100
M.be <- 1
eps.be <- 0.001

mu.prop.var    <- 0.5
sigSq.prop.var    <- 0.01

##

mcmcParams <- list(run=list(numReps=numReps, thin=thin, burninPerc=burninPerc),
tuning=list(mu.prop.var=mu.prop.var, sigSq.prop.var=sigSq.prop.var,
L.beC=L.beC, M.beC=M.beC, eps.beC=eps.beC,
L.be=L.be, M.be=M.be, eps.be=eps.be))

#####################
## Starting Values

w        <- log(Y[,1])
mu     <- 0.1
beta     <- rep(2, p)
sigSq    <- 0.5
tauSq <- rep(0.4, p)
lambdaSq <- 100
betaC     <- rep(0.11, q)

startValues <- list(w=w, beta=beta, tauSq=tauSq, mu=mu, sigSq=sigSq,
lambdaSq=lambdaSq, betaC=betaC)

fit <- aftGL_LT(Y, X, XC, grpInx, hyperParams, startValues, mcmcParams)


## End(Not run)

Function to Fit the Penalized Semiparametric Bayesian Cox Model with Elastic Net Prior

Description

Penalized semiparametric Bayesian Cox (PSBC) model with elastic net prior is implemented to analyze survival data with high-dimensional covariates.

Usage

psbcEN(survObj, priorPara, initial, rw=FALSE, mcmcPara, num.reps, 
		thin, chain = 1, save = 1000)
psbcEN(survObj, priorPara, initial, rw=FALSE, mcmcPara, num.reps, 
		thin, chain = 1, save = 1000)

Arguments

`survObj`	The list containing observed data from `n` subjects; `t`, `di`, `x`
`priorPara`	The list containing prior parameter values; `eta0`, `kappa0`, `c0`, `r1`, `r2`, `delta1`, `delta2`, `s`
`initial`	The list containing the starting values of the parameters; `beta.ini`, `lambda1Sq`, `lambda2`, `sigmaSq`, `tauSq`, `h`
`rw`	When setting to "TRUE", the conventional random walk Metropolis Hastings algorithm is used. Otherwise, the mean and the variance of the proposal density is updated using the jumping rule described in Lee et al. (2011).
`mcmcPara`	The list containing the values of options for Metropolis-Hastings step for $\beta$ ; `numBeta`, `beta.prop.var`
`num.reps`	the number of iterations of the chain
`thin`	thinning
`chain`	the numeric name of chain in the case when running multiple chains.
`save`	frequency of storing the results in .Rdata file. For example, by setting "save = 1000", the algorithm saves the results every 1000 iterations.

Details

`t`	a vector of `n` times to the event
`di`	a vector of `n` censoring indicators for the event time (1=event occurred, 0=censored)
`x`	covariate matrix, `n` observations by `p` variables
`eta0`	scale parameter of gamma process prior for the cumulative baseline hazard, $eta0 > 0$
`kappa0`	shape parameter of gamma process prior for the cumulative baseline hazard, $kappa0 > 0$
`c0`	the confidence parameter of gamma process prior for the cumulative baseline hazard, $c0 > 0$
`r1`	the shape parameter of the gamma prior for $\lambda_1^2$
`r2`	the shape parameter of the gamma prior for $\lambda_2$
`delta1`	the rate parameter of the gamma prior for $\lambda_1^2$
`delta2`	the rate parameter of the gamma prior for $\lambda_2$
`s`	the set of time partitions for specification of the cumulative baseline hazard function
`beta.ini`	the starting values for $\beta$
`lambda1Sq`	the starting value for $\lambda_1^2$
`lambda2`	the starting value for $\lambda_2$
`sigmaSq`	the starting value for $\sigma^2$
`tauSq`	the starting values for $\tau^2$
`h`	the starting values for $h$
`numBeta`	the number of components in $\beta$ to be updated at one iteration
`beta.prop.var`	the variance of the proposal density for $\beta$ when `rw` is set to "TRUE"

Value

psbcEN returns an object of class psbcEN

`beta.p`	posterior samples for $\beta$
`h.p`	posterior samples for $h$
`tauSq.p`	posterior samples for $\tau^2$
`mcmcOutcome`	The list containing posterior samples for the remaining model parameters

Note

If the prespecified value of save is less than that of num.reps, the results are saved as .Rdata file under the directory working directory/mcmcOutcome.

Author(s)

Kyu Ha Lee, Sounak Chakraborty, (Tony) Jianguo Sun

References

Lee, K. H., Chakraborty, S., and Sun, J. (2011). Bayesian Variable Selection in Semiparametric Proportional Hazards Model for High Dimensional Survival Data. The International Journal of Biostatistics, Volume 7, Issue 1, Pages 1-32.

Lee, K. H., Chakraborty, S., and Sun, J. (2015). Survival Prediction and Variable Selection with Simultaneous Shrinkage and Grouping Priors. Statistical Analysis and Data Mining, Volume 8, Issue 2, pages 114-127.

Examples


## Not run: 

# generate some survival data
	
	set.seed(204542)
	
	p = 20
	n = 100
	beta.true <- c(rep(4, 10), rep(0, (p-10)))	
	
	CovX<- diag(0.1, p)
	
	survObj 	<- list()
	survObj$x	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)
	
	pred <- as.vector(exp(rowSums(scale(survObj$x, center = FALSE, scale = 1/beta.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	survObj$t 		<- pmin(t, cen)
	survObj$di 		<- as.numeric(t <= cen)

	priorPara 			<- list()
	priorPara$eta0 		<- 1
	priorPara$kappa0 	<- 1
	priorPara$c0 		<- 2
	priorPara$r1		<- 0.1
	priorPara$r2		<- 1
	priorPara$delta1	<- 0.1
	priorPara$delta2	<- 1
	priorPara$s			<- sort(survObj$t[survObj$di == 1])
	priorPara$s			<- c(priorPara$s, 2*max(survObj$t)
	- max(survObj$t[-which(survObj$t==max(survObj$t))]))
	priorPara$J			<- length(priorPara$s)

	mcmcPara				<- list()
	mcmcPara$numBeta		<- p
	mcmcPara$beta.prop.var	<- 1

	initial				<- list()
	initial$beta.ini	<- rep(0.5, p)
	initial$lambda1Sq	<- 1  
	initial$lambda2		<- 1  
	initial$sigmaSq		<- runif(1, 0.1, 10)
	initial$tauSq		<- rexp(p, rate = initial$lambda1Sq/2)
	initial$h			<- rgamma(priorPara$J, 1, 1)

	rw = FALSE
	num.reps = 20000
	chain = 1
	thin = 5
	save = 5

	fitEN <- psbcEN(survObj, priorPara, initial, rw=FALSE, mcmcPara, 
				num.reps, thin, chain, save)

	vs <- VS(fitEN, X=survObj$x)
    
	
## End(Not run)		

## Not run: 

# generate some survival data
	
	set.seed(204542)
	
	p = 20
	n = 100
	beta.true <- c(rep(4, 10), rep(0, (p-10)))	
	
	CovX<- diag(0.1, p)
	
	survObj 	<- list()
	survObj$x	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)
	
	pred <- as.vector(exp(rowSums(scale(survObj$x, center = FALSE, scale = 1/beta.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	survObj$t 		<- pmin(t, cen)
	survObj$di 		<- as.numeric(t <= cen)

	priorPara 			<- list()
	priorPara$eta0 		<- 1
	priorPara$kappa0 	<- 1
	priorPara$c0 		<- 2
	priorPara$r1		<- 0.1
	priorPara$r2		<- 1
	priorPara$delta1	<- 0.1
	priorPara$delta2	<- 1
	priorPara$s			<- sort(survObj$t[survObj$di == 1])
	priorPara$s			<- c(priorPara$s, 2*max(survObj$t)
	- max(survObj$t[-which(survObj$t==max(survObj$t))]))
	priorPara$J			<- length(priorPara$s)

	mcmcPara				<- list()
	mcmcPara$numBeta		<- p
	mcmcPara$beta.prop.var	<- 1

	initial				<- list()
	initial$beta.ini	<- rep(0.5, p)
	initial$lambda1Sq	<- 1  
	initial$lambda2		<- 1  
	initial$sigmaSq		<- runif(1, 0.1, 10)
	initial$tauSq		<- rexp(p, rate = initial$lambda1Sq/2)
	initial$h			<- rgamma(priorPara$J, 1, 1)

	rw = FALSE
	num.reps = 20000
	chain = 1
	thin = 5
	save = 5

	fitEN <- psbcEN(survObj, priorPara, initial, rw=FALSE, mcmcPara, 
				num.reps, thin, chain, save)

	vs <- VS(fitEN, X=survObj$x)
    
	
## End(Not run)

Function to Fit the Penalized Semiparametric Bayesian Cox Model with Fused Lasso Prior

Description

Penalized semiparametric Bayesian Cox (PSBC) model with fused lasso prior is implemented to analyze survival data with high-dimensional covariates.

Usage

psbcFL(survObj, priorPara, initial, rw=FALSE, mcmcPara, num.reps, 
		thin, chain = 1, save = 1000)
psbcFL(survObj, priorPara, initial, rw=FALSE, mcmcPara, num.reps, 
		thin, chain = 1, save = 1000)

Arguments

`survObj`	The list containing observed data from `n` subjects; `t`, `di`, `x`
`priorPara`	The list containing prior parameter values; `eta0`, `kappa0`, `c0`, `r1`, `r2`, `delta1`, `delta2`, `s`
`initial`	The list containing the starting values of the parameters; `beta.ini`, `lambda1Sq`, `lambda2Sq`, `sigmaSq`, `tauSq`, `h`, `wSq`
`rw`	When setting to "TRUE", the conventional random walk Metropolis Hastings algorithm is used. Otherwise, the mean and the variance of the proposal density is updated using the jumping rule described in Lee et al. (2011).
`mcmcPara`	The list containing the values of options for Metropolis-Hastings step for $\beta$ ; `numBeta`, `beta.prop.var`
`num.reps`	the number of iterations of the chain
`thin`	thinning
`chain`	the numeric name of chain in the case when running multiple chains.
`save`	frequency of storing the results in .Rdata file. For example, by setting "save = 1000", the algorithm saves the results every 1000 iterations.

Details

`t`	a vector of `n` times to the event
`di`	a vector of `n` censoring indicators for the event time (1=event occurred, 0=censored)
`x`	covariate matrix, `n` observations by `p` variables
`eta0`	scale parameter of gamma process prior for the cumulative baseline hazard, $eta0 > 0$
`kappa0`	shape parameter of gamma process prior for the cumulative baseline hazard, $kappa0 > 0$
`c0`	the confidence parameter of gamma process prior for the cumulative baseline hazard, $c0 > 0$
`r1`	the shape parameter of the gamma prior for $\lambda_1^2$
`r2`	the shape parameter of the gamma prior for $\lambda_2^2$
`delta1`	the rate parameter of the gamma prior for $\lambda_1^2$
`delta2`	the rate parameter of the gamma prior for $\lambda_2^2$
`s`	the set of time partitions for specification of the cumulative baseline hazard function
`beta.ini`	the starting values for $\beta$
`lambda1Sq`	the starting value for $\lambda_1^2$
`lambda2Sq`	the starting value for $\lambda_2^2$
`sigmaSq`	the starting value for $\sigma^2$
`tauSq`	the starting values for $\tau^2$
`h`	the starting values for $h$
`wSq`	the starting values for $w^2$
`numBeta`	the number of components in $\beta$ to be updated at one iteration
`beta.prop.var`	the variance of the proposal density for $\beta$ when `rw` is set to "TRUE"

Value

psbcFL returns an object of class psbcFL

`beta.p`	posterior samples for $\beta$
`h.p`	posterior samples for $h$
`tauSq.p`	posterior samples for $\tau^2$
`mcmcOutcome`	The list containing posterior samples for the remaining model parameters

Note

If the prespecified value of save is less than that of num.reps, the results are saved as .Rdata file under the directory working directory/mcmcOutcome.

Author(s)

Kyu Ha Lee, Sounak Chakraborty, (Tony) Jianguo Sun

References

Examples


## Not run: 

# generate some survival data
	
	set.seed(204542)
	
	p = 20
	n = 100
	beta.true <- c(rep(4, 10), rep(0, (p-10)))	
	
	CovX<- diag(0.1, p)
	
	survObj 	<- list()
	survObj$x	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)
	
	pred <- as.vector(exp(rowSums(scale(survObj$x, center = FALSE, scale = 1/beta.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	survObj$t 		<- pmin(t, cen)
	survObj$di 		<- as.numeric(t <= cen)

	priorPara 			<- list()
	priorPara$eta0 		<- 2
	priorPara$kappa0 	<- 2
	priorPara$c0 		<- 2
	priorPara$r1		<- 0.5
	priorPara$r2		<- 0.5
	priorPara$delta1	<- 0.0001
	priorPara$delta2	<- 0.0001
	priorPara$s			<- sort(survObj$t[survObj$di == 1])
	priorPara$s			<- c(priorPara$s, 2*max(survObj$t)
	-max(survObj$t[-which(survObj$t==max(survObj$t))]))
	priorPara$J			<- length(priorPara$s)

	mcmcPara				<- list()
	mcmcPara$numBeta		<- p
	mcmcPara$beta.prop.var	<- 1

	initial				<- list()
	initial$beta.ini	<- rep(0.5, p)
	initial$lambda1Sq	<- 1  
	initial$lambda2Sq	<- 1  
	initial$sigmaSq		<- runif(1, 0.1, 10)
	initial$tauSq		<- rexp(p, rate = initial$lambda1Sq/2)
	initial$h			<- rgamma(priorPara$J, 1, 1)
	initial$wSq	 		<- rexp((p-1), rate = initial$lambda2Sq/2)

	rw = FALSE
	num.reps = 20000
	chain = 1
	thin = 5
	save = 5

	fitFL <- psbcFL(survObj, priorPara, initial, rw=FALSE, mcmcPara, 
				num.reps, thin, chain, save)
	vs <- VS(fitFL, X=survObj$x)
    
	
## End(Not run)						

## Not run: 

# generate some survival data
	
	set.seed(204542)
	
	p = 20
	n = 100
	beta.true <- c(rep(4, 10), rep(0, (p-10)))	
	
	CovX<- diag(0.1, p)
	
	survObj 	<- list()
	survObj$x	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)
	
	pred <- as.vector(exp(rowSums(scale(survObj$x, center = FALSE, scale = 1/beta.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	survObj$t 		<- pmin(t, cen)
	survObj$di 		<- as.numeric(t <= cen)

	priorPara 			<- list()
	priorPara$eta0 		<- 2
	priorPara$kappa0 	<- 2
	priorPara$c0 		<- 2
	priorPara$r1		<- 0.5
	priorPara$r2		<- 0.5
	priorPara$delta1	<- 0.0001
	priorPara$delta2	<- 0.0001
	priorPara$s			<- sort(survObj$t[survObj$di == 1])
	priorPara$s			<- c(priorPara$s, 2*max(survObj$t)
	-max(survObj$t[-which(survObj$t==max(survObj$t))]))
	priorPara$J			<- length(priorPara$s)

	mcmcPara				<- list()
	mcmcPara$numBeta		<- p
	mcmcPara$beta.prop.var	<- 1

	initial				<- list()
	initial$beta.ini	<- rep(0.5, p)
	initial$lambda1Sq	<- 1  
	initial$lambda2Sq	<- 1  
	initial$sigmaSq		<- runif(1, 0.1, 10)
	initial$tauSq		<- rexp(p, rate = initial$lambda1Sq/2)
	initial$h			<- rgamma(priorPara$J, 1, 1)
	initial$wSq	 		<- rexp((p-1), rate = initial$lambda2Sq/2)

	rw = FALSE
	num.reps = 20000
	chain = 1
	thin = 5
	save = 5

	fitFL <- psbcFL(survObj, priorPara, initial, rw=FALSE, mcmcPara, 
				num.reps, thin, chain, save)
	vs <- VS(fitFL, X=survObj$x)
    
	
## End(Not run)

Function to Fit the Penalized Semiparametric Bayesian Cox Model with Group Lasso Prior

Description

Penalized semiparametric Bayesian Cox (PSBC) model with group lasso prior is implemented to analyze survival data with high-dimensional covariates.

Usage

psbcGL(survObj, priorPara, initial, rw=FALSE, mcmcPara, num.reps, 
		thin, chain = 1, save = 1000)
psbcGL(survObj, priorPara, initial, rw=FALSE, mcmcPara, num.reps, 
		thin, chain = 1, save = 1000)

Arguments

`survObj`	The list containing observed data from `n` subjects; `t`, `di`, `x`
`priorPara`	The list containing prior parameter values; `eta0`, `kappa0`, `c0`, `r`, `delta`, `s`, `groupInd`
`initial`	The list containing the starting values of the parameters; `beta.ini`, `lambdaSq`, `sigmaSq`, `tauSq`, `h`
`rw`	When setting to "TRUE", the conventional random walk Metropolis Hastings algorithm is used. Otherwise, the mean and the variance of the proposal density is updated using the jumping rule described in Lee et al. (2011).
`mcmcPara`	The list containing the values of options for Metropolis-Hastings step for $\beta$ ; `numBeta`, `beta.prop.var`
`num.reps`	the number of iterations of the chain
`thin`	thinning
`chain`	the numeric name of chain in the case when running multiple chains.
`save`	frequency of storing the results in .Rdata file. For example, by setting "save = 1000", the algorithm saves the results every 1000 iterations.

Details

`t`	a vector of `n` times to the event
`di`	a vector of `n` censoring indicators for the event time (1=event occurred, 0=censored)
`x`	covariate matrix, `n` observations by `p` variables
`eta0`	scale parameter of gamma process prior for the cumulative baseline hazard, $eta0 > 0$
`kappa0`	shape parameter of gamma process prior for the cumulative baseline hazard, $kappa0 > 0$
`c0`	the confidence parameter of gamma process prior for the cumulative baseline hazard, $c0 > 0$
`r`	the shape parameter of the gamma prior for $\lambda^2$
`delta`	the rate parameter of the gamma prior for $\lambda^2$
`s`	the set of time partitions for specification of the cumulative baseline hazard function
`groupInd`	a vector of `p` group indicator for each variable
`beta.ini`	the starting values for $\beta$
`lambdaSq`	the starting value for $\lambda^2$
`sigmaSq`	the starting value for $\sigma^2$
`tauSq`	the starting values for $\tau^2$
`h`	the starting values for $h$
`numBeta`	the number of components in $\beta$ to be updated at one iteration
`beta.prop.var`	the variance of the proposal density for $\beta$ when `rw` is set to "TRUE"

Value

psbcGL returns an object of class psbcGL

`beta.p`	posterior samples for $\beta$
`h.p`	posterior samples for $h$
`tauSq.p`	posterior samples for $\tau^2$
`mcmcOutcome`	The list containing posterior samples for the remaining model parameters

Note

To fit the PSBC model with the ordinary Bayesian lasso prior (Lee et al., 2011), groupInd needs to be set to 1:p. If the prespecified value of save is less than that of num.reps, the results are saved as .Rdata file under the directory working directory/mcmcOutcome.

Author(s)

Kyu Ha Lee, Sounak Chakraborty, (Tony) Jianguo Sun

References

Examples


## Not run: 

# generate some survival data
	
	set.seed(204542)
	
	p = 20
	n = 100
	beta.true <- c(rep(4, 10), rep(0, (p-10)))	

	CovX<-matrix(0,p,p)

	for(i in 1:10){
		for(j in 1:10){
			CovX[i,j] <- 0.5^abs(i-j)
			}
		}
		
	diag(CovX) <- 1
	
	survObj 	<- list()
	survObj$x	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)
	
	pred <- as.vector(exp(rowSums(scale(survObj$x, center = FALSE, scale = 1/beta.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	survObj$t 		<- pmin(t, cen)
	survObj$di 		<- as.numeric(t <= cen)

	priorPara 			<- list()
	priorPara$eta0 		<- 1
	priorPara$kappa0 	<- 1
	priorPara$c0 		<- 2
	priorPara$r			<- 0.5
	priorPara$delta		<- 0.0001
	priorPara$s			<- sort(survObj$t[survObj$di == 1])
	priorPara$s			<- c(priorPara$s, 2*max(survObj$t)
	-max(survObj$t[-which(survObj$t==max(survObj$t))]))
	priorPara$J			<- length(priorPara$s)
	priorPara$groupInd	<- c(rep(1,10),2:11)

	mcmcPara				<- list()
	mcmcPara$numBeta		<- p
	mcmcPara$beta.prop.var	<- 1

	initial				<- list()
	initial$beta.ini	<- rep(0.5, p)
	initial$lambdaSq	<- 1
	initial$sigmaSq		<- runif(1, 0.1, 10)
	initial$tauSq		<- rexp(length(unique(priorPara$groupInd)),
	rate = initial$lambdaSq/2)
	initial$h			<- rgamma(priorPara$J, 1, 1)

	rw = FALSE
	num.reps = 20000
	chain = 1
	thin = 5
	save = 5

	fitGL <- psbcGL(survObj, priorPara, initial, rw=FALSE, mcmcPara, 
				num.reps, thin, chain, save)
	vs <- VS(fitGL, X=survObj$x)
				
	
## End(Not run)		

## Not run: 

# generate some survival data
	
	set.seed(204542)
	
	p = 20
	n = 100
	beta.true <- c(rep(4, 10), rep(0, (p-10)))	

	CovX<-matrix(0,p,p)

	for(i in 1:10){
		for(j in 1:10){
			CovX[i,j] <- 0.5^abs(i-j)
			}
		}
		
	diag(CovX) <- 1
	
	survObj 	<- list()
	survObj$x	<- apply(rmvnorm(n, sigma=CovX, method="chol"), 2, scale)
	
	pred <- as.vector(exp(rowSums(scale(survObj$x, center = FALSE, scale = 1/beta.true))))
	
	t 		<- rexp(n, rate = pred)
	cen		<- runif(n, 0, 8)      
	survObj$t 		<- pmin(t, cen)
	survObj$di 		<- as.numeric(t <= cen)

	priorPara 			<- list()
	priorPara$eta0 		<- 1
	priorPara$kappa0 	<- 1
	priorPara$c0 		<- 2
	priorPara$r			<- 0.5
	priorPara$delta		<- 0.0001
	priorPara$s			<- sort(survObj$t[survObj$di == 1])
	priorPara$s			<- c(priorPara$s, 2*max(survObj$t)
	-max(survObj$t[-which(survObj$t==max(survObj$t))]))
	priorPara$J			<- length(priorPara$s)
	priorPara$groupInd	<- c(rep(1,10),2:11)

	mcmcPara				<- list()
	mcmcPara$numBeta		<- p
	mcmcPara$beta.prop.var	<- 1

	initial				<- list()
	initial$beta.ini	<- rep(0.5, p)
	initial$lambdaSq	<- 1
	initial$sigmaSq		<- runif(1, 0.1, 10)
	initial$tauSq		<- rexp(length(unique(priorPara$groupInd)),
	rate = initial$lambdaSq/2)
	initial$h			<- rgamma(priorPara$J, 1, 1)

	rw = FALSE
	num.reps = 20000
	chain = 1
	thin = 5
	save = 5

	fitGL <- psbcGL(survObj, priorPara, initial, rw=FALSE, mcmcPara, 
				num.reps, thin, chain, save)
	vs <- VS(fitGL, X=survObj$x)
				
	
## End(Not run)

Penalized Parametric and Semiparametric Bayesian Survival Models with Shrinkage and Grouping Priors

Description

The package provides algorithms for fitting penalized parametric and semiparametric Bayesian survival models with elastic net, fused lasso, and group lasso priors.

Details

The package includes following functions:

`psbcEN`	The function to fit the PSBC model with elastic net prior
`psbcFL`	The function to fit the PSBC model with fused lasso prior
`psbcGL`	The function to fit the PSBC model with group lasso or Bayesian lasso prior
`aftGL`	The function to fit the parametric accelerated failure time model with group lasso
`aftGL_LT`	The function to fit the parametric accelerated failure time model with group lasso for left-truncated and interval-censored data

Package:	psbcGroup
Type:	Package
Version:	1.7
Date:	2024-1-9
License:	GPL (>= 2)
LazyLoad:	yes

Author(s)

Kyu Ha Lee, Sounak Chakraborty, Harrison Reeder, (Tony) Jianguo Sun
Maintainer: Kyu Ha Lee <[email protected]>

References

Reeder, H., Haneuse, S., Lee, K. H. (2023+). Group Lasso Priors for Bayesian Accelerated Failure Time Models with Left-Truncated and Interval-Censored Time-to-Event Data. under review

A simulated survival dataset.

Description

Univariate survival data.

Usage

data(survData)data(survData)

Format

a data frame with 2000 observations on the following 4 variables.

time: the time to event
event: the censoring indicators for the event time; 1=event observed, 0=censored
cluster: cluster numbers
cov1: the first column of covariate matrix x
cov2: the second column of covariate matrix x

Examples

data(survData)
data(survData)

Function to perform variable selection using SNC-BIC thresholding method

Description

The VS is a function to perform variable selection using SNC-BIC thresholding method

Usage

VS(fit, X, psiVec=seq(0.001, 1, 0.001))
VS(fit, X, psiVec=seq(0.001, 1, 0.001))

Arguments

`fit`	an object of class `aftGL`, `psbcEN`, `psbcFL`, or `psbcGL`.
`X`	a covariate matrix, `n` observations by `p` variables
`psiVec`	a vector of candidate threshold values for the SNC step

Author(s)

Kyu Ha Lee

Package 'psbcGroup'

Help Index

Function to Fit the Penalized Parametric Bayesian Accelerated Failure Time Model with Group Lasso Prior

Description

Usage

Arguments

Value

Author(s)

References

See Also

Examples

Function to Fit the Penalized Parametric Bayesian Accelerated Failure Time Model with Group Lasso Prior for Left-Truncated and Interval-Censored Data

Description

Usage

Arguments

Value

Author(s)

References

See Also

Examples

Function to Fit the Penalized Semiparametric Bayesian Cox Model with Elastic Net Prior

Description

Usage

Arguments

Details

Value

Note

Author(s)

References

Examples

Function to Fit the Penalized Semiparametric Bayesian Cox Model with Fused Lasso Prior

Description

Usage

Arguments

Details

Value

Note

Author(s)

References

Examples

Function to Fit the Penalized Semiparametric Bayesian Cox Model with Group Lasso Prior

Description

Usage

Arguments

Details

Value

Note

Author(s)

References

Examples

Penalized Parametric and Semiparametric Bayesian Survival Models with Shrinkage and Grouping Priors

Description

Details

Author(s)

References

A simulated survival dataset.

Description

Usage

Format

Examples

Function to perform variable selection using SNC-BIC thresholding method

Description

Usage

Arguments

Author(s)

References

See Also