bayesmetab: bayesmetab

In nickbond/BASEmetab: Estimate Single-station Whole-stream Metabolic Rates from Diel Dissolved Oxygen (DO) Curves

Description

Estimates single-station whole-stream metabolic rates from diel dissolved oxygen (DO) curves (see Grace et al. 2015).

Usage

bayesmetab(data.dir, results.dir, interval, n.iter = 20000,
  n.burnin = n.iter * 0.5, K.init = 2, smooth.DO = 0,
  smooth.PAR = FALSE, instant = FALSE, update.chains = TRUE,
  extra.iter = 1, K.est = TRUE, K.meas.mean = 0, K.meas.sd = 4,
  p.est = FALSE, theta.est = FALSE)

Arguments

data.dir	relative or absolute path to the folder containing csv input data files to be read.
results.dir	relative or absolute path to the output folder where results (plots and tables) will be written.
interval	Integer. The time interval in seconds (e.g. 10 minutes = 600 seconds)
n.iter	Integer. Number of MCMC iterations (default = 20000)
n.burnin	Integer. Number of iterations of MCMC chains to delete
K.init	Numeric. Initial value of chains for K (/day). Reasonable estimate aids convergence. (default value = 2)
smooth.DO	Numeric. Proportion of high-frequency fluctuations to filter with fast Fourier transform (default = 0)
smooth.PAR	Logical. Should PAR be smoothed with a moving average? (default = FALSE)
instant	Logical. Should a table of instantaneous rates be written? (default = FALSE)
update.chains	Logical. Should the chains automatically update once if not converged? (default = TRUE)
extra.iter	Numeric. Number of extra iterations to run if chains are not converged, as multiple of n.iter (default = 1 times)
K.est	Logical. Should K be estimated with uninformative priors? (default = TRUE)
K.meas.mean	Numeric. Mean for informed normal prior distribution when K.est = FALSE
K.meas.sd	Numeric. Standard deviation for informed normal prior distribution when K.est = FALSE
p.est	Logical. Should p be estimated? (default = FALSE)
theta.est	Logical. Should theta be estimated? (default = FALSE)

Value

A dataframe and csv file of parameter estimates (mean, SD) and checks of model fit, plots of model fit (see Vignette for details https://github.com/dgiling/BASEmetab/blob/master/vignettes/BASEmetab.pdf).

Author(s)

Darren Giling, Ralph Mac Nally

References

Grace et al. (2015) Fast processing of diel oxygen curves: estimating stream metabolism with BASE (BAyesian Single-station Estimation). Limnology and Oceanography: Methods, 13, 103-114.

Examples

##Link to JAGS
library(R2jags)

##View example data set.
#set path to example data.
data.dir <- system.file("extdata", package = "BASEmetab")
ex.data <- read.csv(file.path(data.dir, "Yallakool_example.csv"))
head(ex.data)
tail(ex.data)

##Run Example.

#set output directory to Output folder in current working directory.
results.dir <- file.path(getwd(), "Output")
if (dir.exists(results.dir)){} else {
dir.create(results.dir)}

#run model.
results <- bayesmetab(data.dir, results.dir, interval=600)

nickbond/BASEmetab documentation built on May 28, 2019, 9:31 a.m.