Produce a scatterplot annotated by the correlation, p-value, and regression line.
verboseScatterplot(x, y,
sample = NULL,
corFnc = "cor", corOptions = "use = 'p'",
main = "", xlab = NA, ylab = NA,
cex = 1, cex.axis = 1.5, cex.lab = 1.5, cex.main = 1.5,
abline = FALSE, abline.color = 1, abline.lty = 1,
corLabel = corFnc,
displayAsZero = 1e-5,
col = 1, bg = 0, pch = 1,
lmFnc = lm,
plotPriority = NULL,
showPValue = TRUE,
...)
If sample
above is given, the indices of the plotted points are returned invisibly.
numerical vector to be plotted along the x axis.
numerical vector to be plotted along the y axis.
determines whether x
and y
should be sampled for plotting, useful
to keep the plot manageable when x
and y
are large vectors.
The default NULL
value implies no sampling. A single numeric value will be interpreted as the
number of points to sample randomly. If a vector is given, it will be interpreted as the indices of the
entries in x
and y
that should be plotted. In either case, the correlation and p value will
be determined from the full vectors x
and y
.
character string giving the correlation function to annotate the plot.
character string giving further options to the correlation function.
main title for the plot.
label for the x-axis.
label for the y-axis.
character expansion factor for plot annotations, recycled as necessary.
character expansion factor for axis annotations.
character expansion factor for axis labels.
character expansion factor for the main title.
logical: should the linear regression fit line be plotted?
color specification for the fit line.
line type for the fit line.
character string to be used as the label for the correlation value printed in the main title.
Correlations whose absolute value is smaller than this number will be displayed as zero. This can result in a more intuitive display (for example, cor=0 instead of cor=2.6e-17).
color of the plotted symbols. Recycled as necessary.
fill color of the plotted symbols (used for certain symbols). Recycled as necessary.
Integer code for plotted symbols (see link{plot.default}
). Recycled as necessary.
linear model fit function. Used to calculate the linear model fit line if 'abline'
is
TRUE
. For example, robust linear models are implemented in the
function rlm
.
Optional numeric vector of same length as x
. Points with higher plot priority
will be plotted later, making them more visible if points overlap.
Logical: should the p-value corresponding to the correlation be added to the title?
other arguments to the function plot
.
Steve Horvath and Peter Langfelder
Irrespective of the specified correlation function, the p-value is always calculated for pearson correlation.
plot.default
for standard scatterplots