calib.t: Comparison dates calibrated using both the t distribution...
In Maarten14C/clam: Classical Age-Depth Modelling of Cores from Deposits

View source: R/calibrate.R

calib.t

R Documentation

Comparison dates calibrated using both the t distribution (Christen and Perez 2009) and the normal distribution.

Description

Visualise how a date calibrates using the t distribution and the normal distribution.

Usage

calib.t(
  y = 2450,
  error = 50,
  t.a = 3,
  t.b = 4,
  cc = 1,
  postbomb = NULL,
  cc1 = "IntCal20",
  cc2 = "Marine20",
  cc3 = "SHCal20",
  cc4 = "mixed",
  ccdir = "",
  Cutoff = 1e-05,
  times = 8,
  rule = 1
)

Arguments

`y`	The reported mean of the date.
`error`	The reported error of the date.
`t.a`	Value for the t parameter `a`.
`t.b`	Value for the t parameter `b`.
`cc`	calibration curve for C14 dates (1, 2 or 3).
`postbomb`	Which postbomb curve to use for negative 14C dates
`cc1`	For northern hemisphere terrestrial C14 dates.
`cc2`	For marine C14 dates.
`cc3`	For southern hemisphere C14 dates.
`cc4`	A custom calibration curve
`ccdir`	Directory where the calibration curves for C14 dates `cc` are allocated. By default `ccdir=""`. Use `ccdir="."` to choose current working directory. Use `ccdir="Curves/"` to choose sub-folder `Curves/`.
`Cutoff`	Threshold above which calibrated probabilities are plotted
`times`	8 by default.
`rule`	How should R's approx function deal with extrapolation. If `rule=1`, the default, then NAs are returned for such points and if it is 2, the value at the closest data extreme is used.

Details

Radiocarbon and other dates are usually modelled using the normal distribution (red curve). The t approach (grey distribution) however allows for wider tails and thus tends to better accommodate outlying dates. This distribution requires two parameters, called 'a' and 'b'.