This function plots a ROC curve. It can accept many arguments to
tweak the appearance of the plot. Two syntaxes are possible: one
object of class “roc”, or either two vectors (response, predictor) or a
formula (response~predictor) as in the roc function.
# S3 method for roc
plot(x, ...)
# S3 method for smooth.roc
plot(x, ...)
# S3 method for roc
plot.roc(x, add=FALSE, reuse.auc=TRUE,
axes=TRUE, legacy.axes=FALSE,
# Generic arguments for par:
xlim=if(x$percent){c(100, 0)} else{c(1, 0)},
ylim=if(x$percent){c(0, 100)} else{c(0, 1)},
xlab=ifelse(x$percent, ifelse(legacy.axes, "100 - Specificity (%)", "Specificity (%)"),
ifelse(legacy.axes, "1 - Specificity", "Specificity")),
ylab=ifelse(x$percent, "Sensitivity (%)", "Sensitivity"),
asp=1,
mar=c(4, 4, 2, 2)+.1,
mgp=c(2.5, 1, 0),
# col, lty and lwd for the ROC line only
col=par("col"),
lty=par("lty"),
lwd=2,
type="l",
# Identity line
identity=!add,
identity.col="darkgrey",
identity.lty=1,
identity.lwd=1,
# Print the thresholds on the plot
print.thres=FALSE,
print.thres.pch=20,
print.thres.adj=c(-.05,1.25),
print.thres.col="black",
print.thres.pattern=ifelse(x$percent, "%.1f (%.1f%%, %.1f%%)", "%.3f (%.3f, %.3f)"),
print.thres.cex=par("cex"),
print.thres.pattern.cex=print.thres.cex,
print.thres.best.method=NULL,
print.thres.best.weights=c(1, 0.5),
# Print the AUC on the plot
print.auc=FALSE,
print.auc.pattern=NULL,
print.auc.x=ifelse(x$percent, 50, .5),
print.auc.y=ifelse(x$percent, 50, .5),
print.auc.adj=c(0,1),
print.auc.col=col,
print.auc.cex=par("cex"),
# Grid
grid=FALSE,
grid.v={if(is.logical(grid) && grid[1]==TRUE)
{seq(0, 1, 0.1) * ifelse(x$percent, 100, 1)}
else if(is.numeric(grid))
{seq(0, ifelse(x$percent, 100, 1), grid[1])} else {NULL}},
grid.h={if (length(grid) == 1) {grid.v}
else if (is.logical(grid) && grid[2]==TRUE)
{seq(0, 1, 0.1) * ifelse(x$percent, 100, 1)}
else if(is.numeric(grid))
{seq(0, ifelse(x$percent, 100, 1), grid[2])} else {NULL}},
grid.lty=3,
grid.lwd=1,
grid.col="#DDDDDD",
# Polygon for the AUC
auc.polygon=FALSE,
auc.polygon.col="gainsboro",
auc.polygon.lty=par("lty"),
auc.polygon.density=NULL,
auc.polygon.angle=45,
auc.polygon.border=NULL,
# Polygon for the maximal AUC possible
max.auc.polygon=FALSE,
max.auc.polygon.col="#EEEEEE",
max.auc.polygon.lty=par("lty"),
max.auc.polygon.density=NULL,
max.auc.polygon.angle=45,
max.auc.polygon.border=NULL,
# Confidence interval
ci=!is.null(x$ci),
ci.type=c("bars", "shape", "no"),
ci.col=ifelse(ci.type=="bars", par("fg"), "gainsboro"),
...)
# S3 method for formula
plot.roc(x, data, subset, na.action, ...)
# S3 method for default
plot.roc(x, predictor, ...)
# S3 method for smooth.roc
plot.roc(x, ...)a roc object from the roc function (for plot.roc.roc), a formula (for plot.roc.formula) or a response vector (for plot.roc.default).
arguments for the roc function.
arguments for model.frame
if TRUE, the ROC curve will be added to an existing plot. If FALSE (default), a new plot will be created.
if TRUE (default) and the “roc” object
contains an “auc” field, re-use these specifications for the
plot (specifically print.auc, auc.polygon and
max.auc.polygon arguments). See details.
a logical indicating if the plot axes must be drawn.
a logical indicating if the specificity axis (x
axis) must be plotted as as decreasing “specificity”
(FALSE, the default) or increasing “1 - specificity”
(TRUE) as in most legacy software. This affects only the
axis, not the plot coordinates.
Generic arguments for the
plot. See plot and plot.window for more details. Only
used if add=FALSE.
color, line type and line width for the ROC curve. See par for more details.
type of plotting as in plot.
logical: whether or not the identity line (no discrimination line) must be displayed. Default: only on new plots.
color, line type and line width for the identity line. Used only if identity=TRUE. See par for more details.
Should a selected set of thresholds be displayed on
the ROC curve? FALSE, NULL or “no”: no threshold is
displayed. TRUE or “best”: the threshold with the
highest sum sensitivity + specificity is plotted (this might be more
than one threshold). “all”: all the points of the ROC
curve. “local maximas”: all the local maximas. Numeric
vector: direct definition of the thresholds to display.
Note that on a smoothed ROC curve, only “best” is supported.
the text pattern for the thresholds, as a sprintf format. Three numerics are passed to sprintf: threshold, specificity, sensitivity.
the character expansion factor (cex) for the threshold text pattern. See par for more details.
if
print.thres="best" or print.thres=TRUE, what method must be used to determine which
threshold is the best. See argument best.method and best.weights to
coords for more details.
boolean. Should the numeric value of AUC be printed on the plot?
the text pattern for the AUC, as a sprintf format. If NULL, a reasonable value is computed that takes partial AUC, CI and percent into account. If the CI of the AUC was computed, three numerics are passed to sprintf: AUC, lower CI bound, higher CI bound. Otherwise, only AUC is passed.
x and y position for the printing of the AUC.
the text adjustment, character expansion factor and color for the printing of the AUC. See par for more details.
boolean or numeric vector of length 1 or 2. Should a background grid be added to the plot? Numeric: show a grid with the specified interval between each line; Logical: show the grid or not. Length 1: same values are taken for horizontal and vertical lines. Length 2: grid value for vertical (grid[1]) and horizontal (grid[2]). Note that these values are used to compute grid.v and grid.h. Therefore if you specify a grid.h and grid.v, it will be ignored.
numeric. The x and y values at which a vertical or horizontal line (respectively) must be drawn. NULL if no line must be added.
the line type (lty), line width (lwd) and color (col) of the lines of the grid. See par for more details. Note that you can pass vectors of length 2, in which case it specifies the vertical (1) and horizontal (2) lines.
boolean. Whether or not to display the area as a polygon.
boolean. Whether or not to display the maximal possible area as a polygon.
boolean. Should we plot the confidence intervals?
type and col arguments for
plot.ci. The special value “no” disables the plotting of confidence intervals.
further arguments passed to or from other methods,
especially arguments for roc and plot.roc.roc when calling
plot.roc.default or plot.roc.formula. Note that the
plot argument for roc is not allowed.
Arguments for auc and graphical functions
plot, abline, polygon,
points, text and plot.ci
if applicable.
This function returns a list of class “roc” invisibly. See roc for more details.
For print.auc, auc.polygon and max.auc.polygon
arguments, an AUC specification is
required. By default, the total AUC is plotted, but you may want a
partial AUCs. The specification is defined by:
the “auc” field in the “roc” object if
reuse.auc is set to TRUE (default). It is naturally
inherited from any call to roc and fits most cases.
passing the specification to auc with …
(arguments partial.auc, partial.auc.correct and
partial.auc.focus). In this case, you must ensure either that
the roc object do not contain an auc field (if
you called roc with auc=FALSE), or set
reuse.auc=FALSE.
If reuse.auc=FALSE the auc function will always
be called with … to determine the specification, even if
the “roc” object do contain an auc field.
As well if the “roc” object do not contain an auc
field, the auc function will always be called with
… to determine the specification.
Warning: if the roc object passed to plot.roc contains an auc
field and reuse.auc=TRUE, auc is not called and
arguments such as partial.auc are silently ignored.
This function is typically called from roc when plot=TRUE (not by
default). plot.roc.formula and plot.roc.default are convenience methods
that build the ROC curve (with the roc function) before
calling plot.roc.roc. You can pass them arguments for both
roc and plot.roc.roc. Simply use plot.roc
that will dispatch to the correct method.
The plotting is done in the following order:
A new plot is created if add=FALSE.
The grid is added if grid.v and grid.h are not NULL.
The maximal AUC polygon is added if max.auc.polygon=TRUE.
The CI shape is added if ci=TRUE, ci.type="shape" and x$ci isn't a “ci.auc”.
The AUC polygon is added if auc.polygon=TRUE.
The identity line if identity=TRUE.
The actual ROC line is added.
The CI bars are added if ci=TRUE, ci.type="bars" and x$ci isn't a “ci.auc”.
The selected thresholds are printed if print.thres is TRUE or numeric.
The AUC is printed if print.auc=TRUE.
Graphical functions are called with suppressWarnings.
Xavier Robin, Natacha Turck, Alexandre Hainard, et al. (2011) ``pROC: an open-source package for R and S+ to analyze and compare ROC curves''. BMC Bioinformatics, 7, 77. DOI: 10.1186/1471-2105-12-77.
# NOT RUN {
# Create a few ROC curves:
data(aSAH)
roc.s100b <- roc(aSAH$outcome, aSAH$s100b)
roc.wfns <- roc(aSAH$outcome, aSAH$wfns)
roc.ndka <- roc(aSAH$outcome, aSAH$wfns)
# Simple example:
plot(roc.s100b)
# Add a smoothed ROC:
plot(smooth(roc.s100b), add=TRUE, col="blue")
legend("bottomright", legend=c("Empirical", "Smoothed"),
col=c(par("fg"), "blue"), lwd=2)
# With more options:
plot(roc.s100b, print.auc=TRUE, auc.polygon=TRUE, grid=c(0.1, 0.2),
grid.col=c("green", "red"), max.auc.polygon=TRUE,
auc.polygon.col="lightblue", print.thres=TRUE)
# To plot a different partial AUC, we need to ignore the existing value
# with reuse.auc=FALSE:
plot(roc.s100b, print.auc=TRUE, auc.polygon=TRUE, partial.auc=c(1, 0.8),
partial.auc.focus="se", grid=c(0.1, 0.2), grid.col=c("green", "red"),
max.auc.polygon=TRUE, auc.polygon.col="lightblue",
print.thres=TRUE, print.thres.adj = c(1, -1),
reuse.auc=FALSE)
# Add a second ROC curve to the previous plot:
plot(roc.wfns, add=TRUE)
# Plot some thresholds, add them to the same plot
plot(roc.ndka, print.thres="best", print.thres.best.method="youden")
plot(roc.ndka, print.thres="best", print.thres.best.method="closest.topleft",
add = TRUE)
plot(roc.ndka, print.thres="best", print.thres.best.method="youden",
print.thres.best.weights=c(50, 0.2),
print.thres.adj = c(1.1, 1.25),
add = TRUE)
# }
Run the code above in your browser using DataLab