exchangeSPsurv: exchangeSPsurv

In Nicolas-Schmidt/BayesSPsurv: Bayesian Spatial Split Population Survival Model

Description

Markov Chain Monte Carlo (MCMC) to run Bayesian split population survival model with exchangeable frailties.

Returns a summary of a exchangeSPsurv object via summary.mcmc.

Print method for a exchangeSPsurv x.

Returns a plot of a exchangeSPsurv object via plot.mcmc.

Usage

exchangeSPsurv(
  duration,
  immune,
  Y0,
  LY,
  S,
  data,
  N,
  burn,
  thin,
  w = c(1, 1, 1),
  m = 10,
  ini.beta = 0,
  ini.gamma = 0,
  ini.W = 0,
  ini.V = 0,
  form = c("Weibull", "exponential", "loglog"),
  prop.varV,
  prop.varW,
  id_WV = unique(data[, S])
)

## S3 method for class 'frailtySPsurv'
summary(object, parameter = character(), ...)

## S3 method for class 'frailtySPsurv'
print(x, ...)

## S3 method for class 'frailtySPsurv'
plot(x, parameter = character(), ...)

Arguments

duration	survival stage equation written in a formula of the form Y ~ X1 + X2 + ... where Y is duration until failure or censoring.
immune	split stage equation written in a formula of the form C ~ Z1 + Z2 + ... where C is a binary indicator of immunity.
Y0	the elapsed time since inception until the beginning of time period (t-1).
LY	last observation year (coded as 1; 0 otherwise) due to censoring or failure.
S	spatial information (e.g. district ID) for each observation that matches the spatial matrix row/column information.
data	data.frame.
N	number of MCMC iterations.
burn	burn-in to be discarded.
thin	thinning to prevent from autocorrelation.
w	size of the slice in the slice sampling for (betas, gammas, rho). Write it as a vector. E.g. c(1,1,1).
m	limit on steps in the slice sampling. A vector of values for beta, gamma, rho.
ini.beta	initial value for the parameter vector beta. By default is 0.
ini.gamma	initial value for the parameter vector gamma. By default is 0.
ini.W	initial value for the parameter vector W. By default is 0.
ini.V	initial value for the parameter vector V. By default is 0.
form	type of parametric model (Weibull, Exponential, or Log-Logistic).
prop.varV	proposed variance for V in Metropolis-Hastings.
prop.varW	proposed variance for W in Metropolis-Hastings.
id_WV	vector of type character that modifies the colnames of W and V in the model’s result. By default is unique(data[,S]).
object	an object of class frailtySPsurv, the output of exchangeSPsurv.
parameter	one of five parameters of the spatialSPsurv output. Indicate either "betas," "gammas," "rho", "lambda" or "delta".
...	additional parameter.
x	an object of class frailtySPsurv, the output of exchangeSPsurv.

Value

exchangeSPsurv returns an object of class "frailtySPsurv".

An "exchangeSPsurv" object has the following elements:

betas	matrix, numeric values of the posterior for each variable in the duration equation .
gammas	matrix, numeric values of the posterior for each variable in the immune equation.
rho	vector, numeric values of rho.
lambda	vector, numeric values of lambda.
delta	vector, numeric values of delta.
W	matrix, numeric values of the posterior for Ws.
V	matrix, numeric values of the posterior for Vs.
X	matrix of X's variables.
Z	matrix of Z's variables.
Y	vector of ‘Y’.
Y0	vector of ‘Y0’.
C	vector of ‘C’.
S	vector of ‘S’.
ini.beta	numeric initial values of beta.
ini.gamma	numeric initial values of gamma.
ini.W	numeric initial values of W.
ini.V	numeric initial values of V.
form	character, type of distribution.
call	description for the model to be estimated.

list. Empirical mean, standard deviation and quantiles for each variable.

list. Empirical mean, standard deviation and quantiles for each variable.

Examples

## 1

walter <- spduration::add_duration(Walter_2015_JCR,"renewed_war",
                                   unitID = "ccode", tID = "year",
                                   freq = "year", ongoing = FALSE)

# add S
walter <- spatial_SA(data = walter, var_ccode = "ccode", threshold = 800L)

set.seed(123456)

model <-
    exchangeSPsurv(
           duration  = duration ~ fhcompor1 + lgdpl + comprehensive + victory +
                       instabl + intensityln + ethfrac + unpko,
           immune    = cured ~ fhcompor1 + lgdpl + victory,
           Y0        = 't.0',
           LY        = 'lastyear',
           S         = 'sp_id' ,
           data      = walter[[1]],
           N         = 100,
           burn      = 10,
           thin      = 10,
           w         = c(1,1,1),
           m         = 10,
           form      = "Weibull",
           prop.varV = 1e-05,
           prop.varW = 1e-05,
)

print(model)

summary(model, parameter = "betas")

# plot(model)


## 2

walter   <- spduration::add_duration(Walter_2015_JCR,"renewed_war",
                                   unitID = "ccode", tID = "year",
                                   freq = "year", ongoing = FALSE)

walter$S <- rep(x = 1:length(unique(walter$ccode)), times = rle(walter$ccode)$lengths)
country  <- countrycode::countrycode(unique(walter$ccode),'gwn','iso3c')

set.seed(123456)

model <-
    exchangeSPsurv(
        duration  = duration ~ fhcompor1 + lgdpl + comprehensive + victory +
                    instabl + intensityln + ethfrac + unpko,
        immune    = cured ~ fhcompor1 + lgdpl + victory,
        Y0        = 't.0',
        LY        = 'lastyear',
        S         = 'S' ,
        data      = walter,
        N         = 100,
        burn      = 10,
        thin      = 10,
        w         = c(1,1,1),
        m         = 10,
        form      = "loglog",
        prop.varV = 1e-05,
        prop.varW = 1e-05,
        id_WV     = country
    )

print(model)

Nicolas-Schmidt/BayesSPsurv documentation built on Sept. 17, 2021, 8:51 a.m.