
Bayesian temperature reconstructions with a second predictor
Source:R/rec.ion.bayesian.R
rec.ion.bayesian.RdReconstructs temperatures (in degrees C) from the posterior draws of cal.ion.bayesian(), the measured D47, and the Ion value of each sample.
Arguments
- calModel
Posterior draws from
cal.draws(), a result fromcal.ion.bayesian(), or thestanfitin it.- recData
The reconstruction dataset. It needs the columns
Sample,D47,D47error(1 SD of the D47 value), andIon. AddN(number of replicates) orD47sd(SD of the replicates) to also getSD. The output ofavg.replicates()has both. It also needsIonError(1 SD ofIon) whenuseIonError = TRUE.- ndraws
Number of draws used for each sample.
- useIonError
Whether to account for the error in
Ion.- level
Probability covered by the interval.
- ...
Arguments from earlier versions (
iter,postcalsamples, andMC). They are ignored.
Value
A data.frame with one row per sample. meanTemp is the posterior mean temperature (in degrees C). SE is the standard error of the temperature, propagated from D47error. SD is the SD of the temperature of a single replicate, propagated from the SD of the replicates (D47error x sqrt(N), or D47sd when N is missing). It is NA without replicate information. Both include the uncertainty in the calibration. median, lower, and upper are the posterior median and the limits of the interval, based on SE.
Details
With flat priors on 10^6/T^2 and Ion, the posterior of Ion follows its measurement, and the posterior of 10^6/T^2 for each draw is normal. Both are sampled directly, so no MCMC run is needed. Draws with 10^6/T^2 at or below zero are not physically possible. They are removed with a warning.
The uncertainty includes the calibration parameters, the residual SD of the calibration (sigma), D47error, and IonError when used.
Examples
# \donttest{
ds <- cal.dataset(error = "S1", nobs = 50, seed = 1)
ds$Ion <- rnorm(nrow(ds))
fit <- cal.ion.bayesian(ds, IonError = 0.1, numSavedSteps = 1000)
#> Error in h(simpleError(msg, call)): error in evaluating the argument 'object' in selecting a method for function 'sampling': Boost not found; call install.packages('BH')
recData <- data.frame(Sample = "A", D47 = 0.6, D47error = 0.01, Ion = 0, IonError = 0.1)
rec.ion.bayesian(fit, recData, useIonError = TRUE)
#> Error: object 'fit' not found
# }