paleofire (version 1.2.4)

SEA: Superposed Epoch Analysis

Description

The function performs a Superposed Epoch Analysis on a provided temporal serie. The function uses pfCircular function for the computation of the block bootstrap procedure. The function could be used on both dendrochronological data and other data expressed in CE ages as well on paleoecological series expressed in BP. Proxy data and ages must be spaced evenly but not necessarily using 1 yr time steps.

Usage

SEA(
  x,
  y,
  lag,
  b = NULL,
  conf = c(0.05, 0.95),
  nboot = 1000,
  center = FALSE,
  normalize = FALSE,
  age = "CE"
)

Arguments

x

data frame or matrix with ages and proxy values, younger ages on top.

y

events dates.

lag

lag time used for calculating the SEA.

b

A numeric giving block size, if NULL the optimal block size is given by: b= 2x(-1 /log(p)), where p is the lag one autocorrelation coefficient of the serie (Adams, Mann & Ammann 2003).

conf

confidence intervals for the block bootstrap procedure.

nboot

number of bootstrap replicates.

center

logical, center each epoch by substracting values to each epoch's mean (default = FALSE).

normalize

logical, normalize each epoch by calculating Z-Score (default = FALSE, see Adams, Mann & Ammann 2003)

age

type of ages used in x[,1] either "CE" for Common Era or "BP" for Before Present.

Value

res

A "pfCircular" object with estimated confidence intervals.

See Also

pfCircular

Examples

# NOT RUN {
## Generate some fake data
set.seed(1)
n <- 100 # number of data points
t <- seq(0,4*pi,,100)
a <- 3
b <- 2
c.unif <- runif(n)
amp <- 4

# generate data and calculate "y"
set.seed(1)
y1 <- a*sin(b*t)+c.unif*amp # add uniform error

# SEA applied to fake dendrochronological data in CE
plot(rev(seq(1901,2000,1)), y1, t="l", ylim=range(y1)*c(1.2))
y=c(1923,1948,1972,1995)
points(y,rep(0,length(y)))
x=data.frame(rev(seq(1901,2000,1)),value=y1)
lag=10

#Perform SEA
res=SEA(x, y, lag, b = NULL, conf = c(0.05, 0.95), nboot = 1000, age="CE")
plot(res,xlim=c(-10,10),xlab="lag",ylab="Composite mean")

# SEA applied to fake paleoecological data in BP
plot(seq(-50,49,1), y1, t="l", ylim=range(y1)*c(1.2),xlim=c(50,-50))
y=1950-c(1923,1948,1972,1995)
points(y,rep(0,length(y)))
x=data.frame(seq(-50,49,1),value=y1)
# Perform SEA
res=SEA(x, y, lag, b = NULL, conf = c(0.05, 0.95), nboot = 1000, age="BP")
plot(res,xlim=c(-10,10),xlab="lag",ylab="Composite mean")
# }