skel.plot {dplR}R Documentation

Skeleton Plot

Description

Automatically generates a skeleton plot of tree-ring data.

Usage

skel.plot(rw.vec, yr.vec = NULL, sname = '', filt.weight = 9, dat.out = FALSE,
  master=FALSE, plot=TRUE)

Arguments

rw.vec a vector of a tree-ring chronology or series
yr.vec optional vector giving years for the tree-ring data
sname an optional string of less than 8 characters giving the id for the data
filt.weight filter length for the hanning filter, defaults to 9
dat.out logical flag indicating whether to return a data.frame containing the data.
master logical flag indicating whether to make the plot with the segments inverted to ease pattern matching. If TRUE the segments will be plotted from the top down and the labels on the x axes will be on the bottom.
plot logical flag indicating whether to make a plot.

Details

This makes a skeleton plot - a plot that gives the relative growth for year t relative to years t-1 and t+1. Note that this plots is a standard plot in dendrochronology and typically made by hand for visually crossdating series. This type of plot might be confusing to those not accustomed to visual crossdating. See references for more information.

The skeleton plot is made by calculating departures from high frequency growth for each year by comparing year t to the surrounding three years (t-1,t,t+1). Low frequency variation is removed using a hanning filter. Relative growth is scaled from one to ten but only values greater than three are plotted. This function's primary effect is to create plot with absolute units that can be printed and compared to other plots. Here, anomalous growth is plotted on a 2mm grid and up to 120 years are plotted on a single row with a maximum of 7 rows (840 years). These plots are designed to be plotted on standard paper using an appropriate device, e.g., postscript with defaults or to pdf with plot width and height to accommodate a landscape plot, e.g., width = 10, height = 7.5, paper = 'USr'. These plots are designed to be printable and cut into strips to align long series. Statistical crossdating is possible if the data are output.

Value

This function is invoked primarily for its side effect, which is to produce a plot. If dat.out is TRUE then a data.frame is returned with the years and height of the skeleton plot segments as columns.

Author(s)

Andy Bunn

References

Stokes, M.A. and Smiley, T.L. (1968) An Introduction to Tree-Ring Dating. The University of Arizona Press. ISBN-13: 978-0816516803.

Sheppard, P.R. (2002) Crossdating Tree Rings Using Skeleton Plotting. http://www.ltrr.arizona.edu/skeletonplot/introcrossdate.htm

See Also

device, hanning

Examples


  data(co021)
  x <- co021[,33]
  x.yrs <- as.numeric(rownames(co021))
  x.name <- colnames(co021)[33]
  # On a raw ring width series - undated
  skel.plot(x)
  # On a raw ring width series - dated with names
  skel.plot(x, yr.vec = x.yrs, sname= x.name, master=TRUE)
  ## Not run: 
    # try crossdating
    y <- co021[,11]
    y.yrs <- as.numeric(rownames(co021))
    y.name <- colnames(co021)[11]
    # send to postscript - 3 pages total
    postscript('xdating.examp.ps')
      # 'Master series' with correct celendar dates
      skel.plot(x, yr.vec = x.yrs, sname= x.name, master=TRUE)
      # Undated series, try to align with last plot
      skel.plot(y)
      # Here's the answer...
      skel.plot(y, yr.vec = y.yrs, sname= y.name)
    dev.off()

    # alternatively send to pdf
    pdf('xdating.examp.pdf', width = 10, height = 7.5, paper = 'USr')
      skel.plot(x, yr.vec = x.yrs, sname= x.name, master=TRUE)
      skel.plot(y)
      skel.plot(y, yr.vec = y.yrs, sname= y.name)
    dev.off()
  ## End(Not run)

[Package dplR version 1.1.8 Index]