sns (version 1.2.2)

predict.sns: Sample-based prediction using "sns" Objects

Description

Method for sample-based prediction using the output of sns.run.

Usage

# S3 method for sns
predict(object, fpred
  , nburnin = max(nrow(object)/2, attr(object, "nnr"))
  , end = nrow(object), thin = 1, ...)
# S3 method for predict.sns
summary(object
  , quantiles = c(0.025, 0.5, 0.975)
  , ess.method = c("coda", "ise"), ...)
# S3 method for summary.predict.sns
print(x, ...)

Value

predict.sns produces a matrix with number of rows equal to the length of prediction vector produces by fpred. Its numnber of columns is equal to the number of samples used within the user-specified range, and after thinning (if any). summary.predict.sns produces sample-based prediction mean, standard deviation, quantiles, and effective sample size.

Arguments

object

Object of class "sns" (output of sns.run) or "predict.sns" (output of predict.sns).

fpred

Prediction function, accepting a single value for the state vector and producing a vector of outputs.

nburnin

Number of burn-in iterations discarded for sample-based prediction.

end

Last iteration used in sample-based prediction.

thin

One out of thin iterations within the specified range are used for sample-based prediction.

quantiles

Values for which sample-based quantiles are calculated.

ess.method

Method used for calculating effective sample size. Default is to call effectiveSize from package coda.

x

An object of class "summary.predict.sns".

...

Arguments passed to/from other functions.

Author

Alireza S. Mahani, Asad Hasan, Marshall Jiang, Mansour T.A. Sharabiani

References

Mahani A.S., Hasan A., Jiang M. & Sharabiani M.T.A. (2016). Stochastic Newton Sampler: The R Package sns. Journal of Statistical Software, Code Snippets, 74(2), 1-33. doi:10.18637/jss.v074.c02

See Also

sns.run

Examples

if (FALSE) {

# using RegressionFactory for generating log-likelihood and derivatives
library("RegressionFactory")

loglike.poisson <- function(beta, X, y) {
  regfac.expand.1par(beta, X = X, y = y,
    fbase1 = fbase1.poisson.log)
}

# simulating data
K <- 5
N <- 1000
X <- matrix(runif(N * K, -0.5, +0.5), ncol = K)
beta <- runif(K, -0.5, +0.5)
y <- rpois(N, exp(X %*% beta))

beta.init <- rep(0.0, K)
beta.smp <- sns.run(beta.init, loglike.poisson,
  niter = 1000, nnr = 20, mh.diag = TRUE, X = X, y = y)

# prediction function for mean response
predmean.poisson <- function(beta, Xnew) exp(Xnew %*% beta)
ymean.new <- predict(beta.smp, predmean.poisson,
                     nburnin = 100, Xnew = X)
summary(ymean.new)

# (stochastic) prediction function for response
predsmp.poisson <- function(beta, Xnew)
  rpois(nrow(Xnew), exp(Xnew %*% beta))
ysmp.new <- predict(beta.smp, predsmp.poisson
                    , nburnin = 100, Xnew = X)
summary(ysmp.new)

}