mcmcsa: MCMC for source apportionment

In kralljr/MCMCsa: Performs Bayesian source apportionment for PM data

Description

mcmcsa performs MCMC for source apportionment

Usage

mcmcsa(x, ...)

## Default S3 method:
mcmcsa(dat, lamcon, mdls = NULL, guessvec = NULL,
  burnin = 10000, N = 1e+05, fix = NULL)

Arguments

dat	data frame of daily constituent concentrations with date as first column
lamcon	Matrix of sources by constituents with 1 and 0 for conditions
mdls	data frame of daily MDL values of the same dimensions of data
guessvec	starting values (do not need to provide)
burnin	number of burnin samples
N	number of samples
L	number of sources

Details

This function aims to draw from the posterior distributions of the source profiles, the source contributions/concentrations, the variance of the data and the other parameters. Eventually, this function will incorporate options for concentrations observed below the MDL

Examples

data(nycdat)
data(nycmdl)
data(lamcon)
#fix data totals
pm <- nycdat$PM25
whPM <- which(colnames(nycdat) == "PM25")
nycdat <- nycdat[, -whPM]
whPM <- which(colnames(nycmdl) == "PM25")
nycmdl <- nycmdl[, -whPM]
L <- 4

mcmcsa(nycdat, L, lamcon)

kralljr/MCMCsa documentation built on May 20, 2019, 1:13 p.m.