## File: initseq.Rd

package info (click to toggle)
r-cran-mcmc 0.9-7-1
 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103 \name{initseq} \alias{initseq} \title{Initial Sequence Estimators} \description{ Variance of sample mean of functional of reversible Markov chain using methods of Geyer (1992). } \usage{ initseq(x) } \arguments{ \item{x}{a numeric vector that is a scalar-valued functional of a reversible Markov chain.} } \details{ Let \deqn{\gamma_k = \textrm{cov}(X_i, X_{i + k})}{gamma[k] = cov(x[i], x[i + k])} considered as a function of the lag \eqn{k} be the autocovariance function of the input time series. Define \deqn{\Gamma_k = \gamma_{2 k} + \gamma_{2 k + 1}}{Gamma[k] = gamma[2 k] + gamma[2 k + 1]} the sum of consecutive pairs of autocovariances. Then Theorem 3.1 in Geyer (1992) says that \eqn{\Gamma_k}{Gamma[k]} considered as a function of \eqn{k} is strictly positive, strictly decreasing, and strictly convex, assuming the input time series is a scalar-valued functional of a reversible Markov chain. All of the MCMC done by this package is reversible. This \R function estimates the \dQuote{big gamma} function, \eqn{\Gamma_k}{Gamma[k]} considered as a function of \eqn{k}, subject to three different constraints, (1) nonnegative, (2) nonnegative and nonincreasing, and (3) nonnegative, nonincreasing, and convex. It also estimates the variance in the Markov chain central limit theorem (CLT) \deqn{\gamma_0 + 2 \sum_{k = 1}^\infty \gamma_k = - \gamma_0 + 2 \sum_{k = 0}^\infty \Gamma_k}{- gamma0 + 2 * sum(gamma) = - gamma0 + 2 * sum(Gamma)} \strong{Note:} The batch means provided by \code{\link{metrop}} are also scalar functionals of a reversible Markov chain. Thus these initial sequence estimators applied to the batch means give valid standard errors for the mean of the match means even when the batch length is too short to provide a valid estimate of asymptotic variance. One does, of course, have to multiply the asymptotic variance of the batch means by the batch length to get the asymptotic variance for the unbatched chain. } \value{ a list containing the following components: \item{gamma0}{the scalar \eqn{\gamma_0}{gamma}, the marginal variance of \code{x}.} \item{Gamma.pos}{the vector \eqn{\Gamma}{Gamma}, estimated so as to be nonnegative, where, as always, \R uses one-origin indexing so \code{Gamma.pos} is \eqn{\Gamma_0}{Gamma}.} \item{Gamma.dec}{the vector \eqn{\Gamma}{Gamma}, estimated so as to be nonnegative and nonincreasing, where, as always, \R uses one-origin indexing so \code{Gamma.dec} is \eqn{\Gamma_0}{Gamma}.} \item{Gamma.con}{the vector \eqn{\Gamma}{Gamma}, estimated so as to be nonnegative and nonincreasing and convex, where, as always, \R uses one-origin indexing so \code{Gamma.con} is \eqn{\Gamma_0}{Gamma}.} \item{var.pos}{the scalar \code{- gamma0 + 2 * sum(Gamma.pos)}, which is the asymptotic variance in the Markov chain CLT. Divide by \code{length(x)} to get the approximate variance of the sample mean of \code{x}.} \item{var.dec}{the scalar \code{- gamma0 + 2 * sum(Gamma.dec)}, which is the asymptotic variance in the Markov chain CLT. Divide by \code{length(x)} to get the approximate variance of the sample mean of \code{x}.} \item{var.con}{the scalar \code{- gamma0 + 2 * sum(Gamma.con)}, which is the asymptotic variance in the Markov chain CLT. Divide by \code{length(x)} to get the approximate variance of the sample mean of \code{x}.} } \section{Bugs}{ Not precisely a bug, but \code{var.pos}, \code{var.dec}, and \code{var.con} can be negative. This happens only when the chain is way too short to estimate the variance, and even then rarely. But it does happen. } \references{ Geyer, C. J. (1992) Practical Markov Chain Monte Carlo. \emph{Statistical Science} \bold{7} 473--483. } \seealso{ \code{\link{metrop}} } \examples{ n <- 2e4 rho <- 0.99 x <- arima.sim(model = list(ar = rho), n = n) out <- initseq(x) \dontrun{ plot(seq(along = out$Gamma.pos) - 1, out$Gamma.pos, xlab = "k", ylab = expression(Gamma[k]), type = "l") lines(seq(along = out$Gamma.dec) - 1, out$Gamma.dec, col = "red") lines(seq(along = out$Gamma.con) - 1, out$Gamma.con, col = "blue") } # asymptotic 95\% confidence interval for mean of x mean(x) + c(-1, 1) * qnorm(0.975) * sqrt(out$var.con / length(x)) # estimated asymptotic variance out$var.con # theoretical asymptotic variance (1 + rho) / (1 - rho) * 1 / (1 - rho^2) # illustrating use with batch means bm <- apply(matrix(x, nrow = 5), 2, mean) initseq(bm)\$var.con * 5 } \keyword{ts}