coreInfluence {Bchron}

Find the influence of dates in a pair of Bchronology runs across the core

Description

This function takes as input two Bchronology runs and compares the uncertainty intervals. It does this by computing the mean uncertainty across the core (type = 'mean') at a specified percentile level (e.g. 95%) and subsequently reporting the reduction/increase in uncertainty between the two runs. Both cores must have the same set of depths/positions at regular intervals.

Usage

coreInfluence(
  bchrRun1,
  bchrRun2,
  percentile = 0.95,
  type = c("plot", "summary", "max"),
  ageTolerance = 500,
  ...
)

Arguments

bchrRun1

The output of a run of the Bchronology function

bchrRun2

The output of another run of the Bchronology function, possibly with different dates. Note this must have the same value of predictPositions as bchrRun1

percentile

The value of the percentile to compare the uncertainties. Default is 95%

type

if plot will return a plot of the difference in uncertainties at the specified percentile level. If summary will return text output of the reduction in uncertainty at each position. If max will return the position of the maximum decrease in uncertainty and a list of all the positions where the reduction in uncertainty exceeds the value of ageTolerance

ageTolerance

A value in years for which to report the positions at which the reduction in uncertainty exceeds this value.

...

Additional arguments to plot

Details

For example, if the ageTolerance value is 500 years, then coreInfluence will return all of the positions at which the uncertainty reduction is bigger than 500.

Value

Depending on type will outputs some text and plots providing the influence values for the cores in question.

See Also

Bchronology, choosePositions, dateInfluence for finding the influence of removing a single date from a core

Examples

data(Glendalough)
# Start with a run that remove two dates
GlenOut1 <- Bchronology(
  ages = Glendalough$ages[-c(3:4)],
  ageSds = Glendalough$ageSds[-c(3:4)],
  calCurves = Glendalough$calCurves[-c(3:4)],
  positions = Glendalough$position[-c(3:4)],
  positionThicknesses = Glendalough$thickness[-c(3:4)],
  ids = Glendalough$id[-c(3:4)],
  predictPositions = seq(0, 1500, by = 10)
)
GlenOut2 <- Bchronology(
  ages = Glendalough$ages,
  ageSds = Glendalough$ageSds,
  calCurves = Glendalough$calCurves,
  positions = Glendalough$position,
  positionThicknesses = Glendalough$thickness,
  ids = Glendalough$id,
  predictPositions = seq(0, 1500, by = 10)
)

# Now compare their influence
coreInfluence(GlenOut1,
  GlenOut2,
  type = c("max", "plot"),
  xlab = "Age (cal years BP)",
  ylab = "Depth (cm)",
  main = "Chronology difference at 95% for
              Glendalough removing two dates",
  las = 1
)