seasonal {tsModel}R Documentation

Reproducible Seasonal Analysis of Air Pollution and Mortality

Description

Functions for conducting a reproducible seasonal analysis of air pollution and mortality in the United States.

Usage

coefSeasonal(results, pollutant, method = "factor2",
             Seasons = c("Winter", "Spring", "Summer", "Fall"))
multiDFFit(dfVec, city, ...)
poolCoef(b, cov = NULL, w = NULL, method = c("tlnise", "fixed"),
         extractors = NULL, ...)
extractBetaCov(results, pollutant)
LouisFormat(x, type = c("stderr", "confint"), digits = 2)
gibbspool(b, v, scale = 1, maxiter = 2000, burn = 500,
          pScale = 1e-5)
pooling(estimate, var)

Arguments

results list of individual city-specific regression fits
pollutant name of pollutant of interest
method type of seasonal model, or method of pooling
Seasons character vector of season names
dfVec vector of degrees of freedom per year to use in the smooth function of time
city name of a city in the NMMAPS database
b a vector of N coefficients or an N x p matrix of coefficient vectors
cov,v a vector of N variances or a p x p x N array of covariances matrices
w second-stage covariates to include in the hierarchical model
extractors a list of functions
x a numeric matrix
type type of output to produce
digits number of digits of output to use
scale prior standard deviation of the national average
maxiter maximum number of iterations in the Gibbs sampler
burn number of iterations to discard
pScale scaling coefficient for the prior variance of the heterogeneity parameter
estimate a vector of coefficients
var a vector of coefficient variances
... other arguments passed to fitCitySeason

Details

These functions are used to reproduce a seasonal analysis of air pollution and mortality in the United States (see the reference below).

Author(s)

Roger D. Peng rpeng@jhsph.edu

References

Peng RD, Dominic F (2008). Statistical Methods for Environmental Epidemiology in R: A Case Study in Air Pollution and Health, Springer.


[Package tsModel version 0.5-1 Index]