plotTaylor {oce} | R Documentation |
Creates a diagram as described by Taylor (2001). The graph is in the form
of a semicircle, with radial lines and spokes connecting at a focus point on
the flat (lower) edge. The radius of a point on the graph indicates the
standard deviation of the corresponding quantity, i.e. x and the columns in
y. The angle connecting a point on the graph to the focus provides an
indication of correlation coefficient with respect to x. The “east” side
of the graph indicates R=1, while R=0 is at the
“north edge” and R=-1 is at the “west” side. The x
data are indicated with a bullet on the graph, appearing on the lower edge
to the right of the focus at a distance indicating the standard deviation of
x
. The other points on the graph represent the columns of y
,
coded automatically or with the supplied values of pch
and
col
.
The example shows two tidal models of the Halifax sealevel data, computed
with tidem
with just the M2 component and the S2 component;
the graph indicates that the M2 model is much better than the S2 model.
plotTaylor(x, y, scale, pch, col, labels, pos, ...)
x |
a vector of reference values of some quantity, e.g. measured over time or space. |
y |
a matrix whose columns hold values of values to be compared with
those in x. (If |
scale |
optional scale, interpreted as the maximum value of standard deviation. |
pch |
list of characters to plot, one for each column of |
col |
list of colours for points on the plot, one for each column of
|
labels |
optional vector of strings to use for labelling the points. |
pos |
optional vector of positions for labelling strings. If not provided, labels will be to the left of the symbols. |
... |
optional arguments passed by |
Dan Kelley
Taylor, Karl E., 2001. Summarizing multiple aspects of model performance in a single diagram, J. Geophys. Res., 106:D7, 7183–7192.
library(oce) data(sealevel) x <- sealevel[["elevation"]] M2 <- predict(tidem(sealevel, constituents="M2")) S2 <- predict(tidem(sealevel, constituents=c("S2"))) plotTaylor(x, cbind(M2, S2))