Type: Package
Title: Generate Useful ROC Curve Charts for Print and Interactive Use
Version: 2.3.1
Date: 2023-10-05
Maintainer: Michael C. Sachs <sachsmc@gmail.com>
Description: Most ROC curve plots obscure the cutoff values and inhibit interpretation and comparison of multiple curves. This attempts to address those shortcomings by providing plotting and interactive tools. Functions are provided to generate an interactive ROC curve plot for web use, and print versions. A Shiny application implementing the functions is also included.
License: MIT + file LICENSE
URL: https://sachsmc.github.io/plotROC/
BugReports: https://github.com/sachsmc/plotROC/issues/
Depends: R (≥ 3.0.0), ggplot2
Imports: methods, grid, gridSVG, shiny, plyr, rlang
Suggests: knitr, testthat, stringr, survivalROC, rmarkdown
VignetteBuilder: knitr
RoxygenNote: 7.2.3
NeedsCompilation: no
Packaged: 2023-10-06 12:13:25 UTC; sachsmc
Author: Michael C. Sachs [aut, cre], Robert W. Corty [ctb], Luis Crouch [ctb] (Modification to calc_auc)
Repository: CRAN
Date/Publication: 2023-10-06 12:40:02 UTC

Tools for plotting ROC Curves

Description

Generate Useful ROC Curve Charts for Print and Interactive Use. This defines a set of stats and geoms for use with ggplot2. In addition, ggplot objects created with these geoms can be exported and turned into interactive plots for use on the web. The interactive features include hovering to display hidden labels, and clicking to reveal confidence regions.

Details

To get started, see geom_roc, geom_rocci, or the examples below. For transforming data, we also provide the convenience function melt_roc.

The vignette contains some examples, along with explanations of the results. To view, run vignette("examples", package = "plotROC")

Author(s)

Michael Sachs (@sachsmc)

See Also

Useful links:

Examples


D.ex <- rbinom(50, 1, .5)
rocdata <- data.frame(D = c(D.ex, D.ex), 
                   M = c(rnorm(50, mean = D.ex, sd = .4), rnorm(50, mean = D.ex, sd = 1)), 
                   Z = c(rep("A", 50), rep("B", 50)))

ggplot(rocdata, aes(m = M, d = D)) + geom_roc() + geom_rocci()

ggplot(rocdata, aes(m = M, d = D)) + geom_roc() + style_roc()
ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc() + geom_rocci()
ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc() + geom_rocci(sig.level = .01)
ggplot(rocdata, aes(m = M, d = D)) + geom_roc(n.cuts = 0) + 
geom_rocci(ci.at = quantile(rocdata$M, c(.1, .25, .5, .75, .9)))
ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc() + geom_rocci(linetype = 1)

rocplot <- ggplot(rocdata, aes(m = M, d = D)) + geom_roc()

plot_interactive_roc(rocplot)
plot_interactive_roc(rocplot + aes(color = Z))
plot_interactive_roc(rocplot + facet_wrap( ~ Z))

Empirical Receiver Operating Characteristic Curve

Description

Display the empirical ROC curve. Useful for characterizing the classification accuracy of continuous measurements for predicting binary states

Usage

GeomRoc

geom_roc(
  mapping = NULL,
  data = NULL,
  stat = "roc",
  n.cuts = 10,
  arrow = NULL,
  lineend = "butt",
  linejoin = "round",
  linemitre = 1,
  linealpha = 1,
  pointalpha = 1,
  pointsize = 0.5,
  labels = TRUE,
  labelsize = 3.88,
  labelround = 1,
  na.rm = TRUE,
  cutoffs.at = NULL,
  cutoff.labels = NULL,
  position = "identity",
  show.legend = NA,
  inherit.aes = TRUE,
  ...
)

Arguments

mapping

Set of aesthetic mappings created by aes(). If specified and inherit.aes = TRUE (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping.

data

The data to be displayed in this layer. There are three options:

If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot().

A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify() for which variables will be created.

A function will be called with a single argument, the plot data. The return value must be a data.frame, and will be used as the layer data. A function can be created from a formula (e.g. ~ head(.x, 10)).

stat

Use to override the default connection between geom_roc and stat_roc.

n.cuts

Number of cutpoints to display along each curve

arrow

Arrow specification, as created by arrow

lineend

Line end style (round, butt, square)

linejoin

Line join style (round, mitre, bevel)

linemitre

Line mitre limit (number greater than 1)

linealpha

Alpha level for the lines, alpha.line is deprecated

pointalpha

Alpha level for the cutoff points, alpha.point is deprecated

pointsize

Size of cutoff points, size.point is deprecated

labels

Logical, display cutoff text labels

labelsize

Size of cutoff text labels

labelround

Integer, number of significant digits to round cutoff labels

na.rm

Remove missing values from curve

cutoffs.at

Vector of user supplied cutoffs to plot as points. If non-NULL, it will override the values of n.cuts and plot the observed cutoffs closest to the user-supplied ones.

cutoff.labels

vector of user-supplied labels for the cutoffs. Must be a character vector of the same length as cutoffs.at.

position

Position adjustment, either as a string naming the adjustment (e.g. "jitter" to use position_jitter), or the result of a call to a position adjustment function. Use the latter if you need to change the settings of the adjustment.

show.legend

logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. It can also be a named logical vector to finely select the aesthetics to display.

inherit.aes

If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders().

...

Other arguments passed on to layer(). These are often aesthetics, used to set an aesthetic to a fixed value, like colour = "red" or size = 3. They may also be parameters to the paired geom/stat.

Format

An object of class GeomRoc (inherits from Geom, ggproto, gg) of length 6.

Computed variables

false_positive_fraction

estimate of false positive fraction

true_positive_fraction

estimate of true positive fraction

cutoffs

values of m at which estimates are calculated

Aesthetics

geom_roc understands the following aesthetics (required aesthetics are in bold):

See Also

See geom_rocci for displaying rectangular confidence regions for the empirical ROC curve, style_roc for adding guidelines and labels, and direct_label for adding direct labels to the curves. Also export_interactive_roc for creating interactive ROC curve plots for use in a web browser.

Examples

D.ex <- rbinom(50, 1, .5)
rocdata <- data.frame(D = c(D.ex, D.ex), 
                   M = c(rnorm(50, mean = D.ex, sd = .4), rnorm(50, mean = D.ex, sd = 1)), 
                   Z = c(rep("A", 50), rep("B", 50)))

ggplot(rocdata, aes(m = M, d = D)) + geom_roc()

ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc()
ggplot(rocdata, aes(m = M, d = D)) + geom_roc() + facet_wrap(~ Z)
ggplot(rocdata, aes(m = M, d = D)) + geom_roc(n.cuts = 20)
ggplot(rocdata, aes(m = M, d = D)) + geom_roc(cutoffs.at = c(1.5, 1, .5, 0, -.5))
ggplot(rocdata, aes(m = M, d = D)) + geom_roc(labels = FALSE)
ggplot(rocdata, aes(m = M, d = D)) + geom_roc(size = 1.25)

Calculate the empirical Receiver Operating Characteristic curve

Description

Given a binary outcome d and continuous measurement m, computes the empirical ROC curve for assessing the classification accuracy of m

Usage

StatRoc

stat_roc(
  mapping = NULL,
  data = NULL,
  geom = "roc",
  position = "identity",
  show.legend = NA,
  inherit.aes = TRUE,
  na.rm = TRUE,
  max.num.points = 1000,
  increasing = TRUE,
  ...
)

Arguments

mapping

Set of aesthetic mappings created by aes(). If specified and inherit.aes = TRUE (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping.

data

The data to be displayed in this layer. There are three options:

If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot().

A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify() for which variables will be created.

A function will be called with a single argument, the plot data. The return value must be a data.frame, and will be used as the layer data. A function can be created from a formula (e.g. ~ head(.x, 10)).

geom

The geometric object to use to display the data, either as a ggproto Geom subclass or as a string naming the geom stripped of the geom_ prefix (e.g. "point" rather than "geom_point")

position

Position adjustment, either as a string naming the adjustment (e.g. "jitter" to use position_jitter), or the result of a call to a position adjustment function. Use the latter if you need to change the settings of the adjustment.

show.legend

logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. It can also be a named logical vector to finely select the aesthetics to display.

inherit.aes

If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders().

na.rm

Remove missing observations

max.num.points

maximum number of points to plot

increasing

TRUE (default) if M is positively associated with Pr(D = 1), if FALSE, assumes M is negatively associated with Pr(D = 1)

...

Other arguments passed on to layer(). These are often aesthetics, used to set an aesthetic to a fixed value, like colour = "red" or size = 3. They may also be parameters to the paired geom/stat.

Format

An object of class StatRoc (inherits from Stat, ggproto, gg) of length 6.

Aesthetics

stat_roc understands the following aesthetics (required aesthetics are in bold):

Computed variables

false_positive_fraction

estimate of false positive fraction

true_positive_fraction

estimate of true positive fraction

cutoffs

values of m at which estimates are calculated

Examples

D.ex <- rbinom(50, 1, .5)
rocdata <- data.frame(D = c(D.ex, D.ex), 
                   M = c(rnorm(50, mean = D.ex, sd = .4), rnorm(50, mean = D.ex, sd = 1)), 
                   Z = c(rep("A", 50), rep("B", 50)))

ggplot(rocdata, aes(m = M, d = D)) + stat_roc()

Calculate confidence regions for the empirical ROC curve

Description

Confidence intervals for TPF and FPF are calculated using the exact method of Clopper and Pearson (1934) each at the level 1 - sqrt(1 - alpha). Based on result 2.4 from Pepe (2003), the cross-product of these intervals yields a 1 - alpha

Usage

StatRocci

stat_rocci(
  mapping = NULL,
  data = NULL,
  geom = "rocci",
  position = "identity",
  show.legend = NA,
  inherit.aes = TRUE,
  ci.at = NULL,
  sig.level = 0.05,
  na.rm = TRUE,
  ...
)

Arguments

mapping

Set of aesthetic mappings created by aes(). If specified and inherit.aes = TRUE (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping.

data

The data to be displayed in this layer. There are three options:

If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot().

A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify() for which variables will be created.

A function will be called with a single argument, the plot data. The return value must be a data.frame, and will be used as the layer data. A function can be created from a formula (e.g. ~ head(.x, 10)).

geom

The geometric object to use to display the data, either as a ggproto Geom subclass or as a string naming the geom stripped of the geom_ prefix (e.g. "point" rather than "geom_point")

position

Position adjustment, either as a string naming the adjustment (e.g. "jitter" to use position_jitter), or the result of a call to a position adjustment function. Use the latter if you need to change the settings of the adjustment.

show.legend

logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. It can also be a named logical vector to finely select the aesthetics to display.

inherit.aes

If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders().

ci.at

Vector of cutoffs at which to display confidence regions. If NULL, will automatically choose 3 evenly spaced points to display the regions

sig.level

Significance level for the confidence regions

na.rm

Remove missing observations

...

Other arguments passed on to layer(). These are often aesthetics, used to set an aesthetic to a fixed value, like colour = "red" or size = 3. They may also be parameters to the paired geom/stat.

Format

An object of class StatRocci (inherits from Stat, ggproto, gg) of length 6.

Aesthetics

stat_rocci understands the following aesthetics (required aesthetics are in bold):

Computed variables

FPF

estimate of false positive fraction

TPF

estimate of true positive fraction

cutoffs

values of m at which estimates are calculated

FPFL

lower bound of confidence region for FPF

FPFU

upper bound of confidence region for FPF

TPFL

lower bound of confidence region for TPF

TPFU

upper bound of confidence region for TPF

References

Examples

D.ex <- rbinom(50, 1, .5)
rocdata <- data.frame(D = c(D.ex, D.ex), 
                   M = c(rnorm(50, mean = D.ex, sd = .4), rnorm(50, mean = D.ex, sd = 1)), 
                   Z = c(rep("A", 50), rep("B", 50)))

ggplot(rocdata, aes(m = M, d = D)) + geom_roc() + stat_rocci()
ggplot(rocdata, aes(m = M, d = D)) + geom_roc() + 
stat_rocci(ci.at = quantile(rocdata$M, c(.1, .3, .5, .7, .9)))

Calculate the Area under the ROC curve

Description

Given a ggplot object with a GeomRoc layer, computes the area under the ROC curve for each group

Usage

calc_auc(ggroc)

Arguments

ggroc

A ggplot object that contains a GeomRoc layer

Value

A data frame with the estimated AUCs for each panel and group

Examples

D.ex <- rbinom(50, 1, .5)
rocdata <- data.frame(D = c(D.ex, D.ex),
                     M = c(rnorm(50, mean = D.ex, sd = .4), rnorm(50, mean = D.ex, sd = 1)),
                     Z = c(rep("A", 50), rep("B", 50)))

ggroc <- ggplot(rocdata, aes(m = M, d = D)) + geom_roc()
calc_auc(ggroc)
ggroc2 <- ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc()
calc_auc(ggroc2)

Calculate the Empirical ROC curves for multiple biomarkers

Description

Deprecated, use geom_roc instead

Usage

calculate_multi_roc(data, M_string, D_string)

Arguments

data

data frame containing at least 1 marker and the common class labels, coded as 0 and 1

M_string

vector of marker column names

D_string

class label column name

Value

List of data frames containing cutoffs, and estimated true and false positive fractions

Calculate the Empirical ROC curve

Description

Deprecated, use geom_roc instead

Usage

calculate_roc(M, D, ci = FALSE, alpha = 0.05)

Arguments

M

continuous marker values or predictions of class labels

D

class labels, must be coded as 0 and 1. If not numeric with 0/1, then plotROC assumes the first level in sort order is healthy status, with a warning.

ci

Logical, if true, will calculate exact joint confidence regions for the TPF and FPF

alpha

Confidence level, ignored if ci = FALSE

Details

Confidence intervals for TPF and FPF are calculated using the exact method of Clopper and Pearson (1934) each at the level 1 - sqrt(1 - alpha). Based on result 2.4 from Pepe (2003), the cross-product of these intervals yields a 1 - alpha

Value

A dataframe containing cutoffs, estimated true and false positive fractions, and confidence intervals if ci = TRUE.

Add direct labels to a ROC plot

Description

Add direct labels to a ROC plot

Usage

direct_label(
  ggroc_p,
  labels = NULL,
  label.angle = 45,
  nudge_x = 0,
  nudge_y = 0,
  size = 6,
  ...
)

Arguments

ggroc_p

A ggplot object that contains a geom_roc layer

labels

vector of labels to add directly to the plot next to the curves. If multiple curves, must be in the same order as the grouping factor. If NULL, attempts to determine labels from the ggroc_p object

label.angle

angle of adjustment for the direct labels

nudge_x, nudge_y

Horizontal and vertical adjustment to nudge labels by. These can be scalars or vectors the same length as the number of labels

size

Size of labels

...

Other arguments passed to annotate

Generate svg code for an ROC curve object

Description

Takes a ggplot object that contains a GeomRoc layer and returns a string that contains html suitable for creating a standalone interactive ROC curve plot.

Usage

export_interactive_roc(
  ggroc_p,
  add.cis = TRUE,
  hide.points = FALSE,
  prefix = "a",
  width = 6,
  height = 6,
  omit.js = FALSE,
  style = style_roc(theme = theme_grey()),
  ...
)

Arguments

ggroc_p

A ggplot object with a GeomRoc layer and optionally a GeomRocci layer as returned by geom_roc and/or geom_rocci. It can be modified with annotations, themes, etc.

add.cis

Logical, if true, removes the current confidence interval layer (if present) and replaces it with a denser layer of confidence regions

hide.points

Logical, if true, hides points layer so that points with cutoff values are only visible when hovering. Recommended for plots containing more than 3 curves.

prefix

A string to assign to the objects within the svg. Enables unique identification by the javascript code

width, height

Width and height in inches of plot

omit.js

Logical. If true, omit inclusion of javascript source in output. Useful for documents with multiple interactive plots

style

A call to the function style_roc

...

Other arguments passed to geom_rocci when add.cis = TRUE

Details

If you intend to include more than one of these objects in a single page, use a different prefix string for each one. To use this function in knitr, use the chunk options fig.keep='none' and results = 'asis', then cat() the resulting string to the output. See the vignette for examples. Older browsers (< IE7) are not supported.

Value

A character object containing the html necessary to plot the ROC curve in a web browser

Confidence regions for the ROC curve

Description

Display rectangular confidence regions for the empirical ROC curve.

Usage

geom_rocci(
  mapping = NULL,
  data = NULL,
  stat = "rocci",
  ci.at = NULL,
  sig.level = 0.05,
  na.rm = TRUE,
  alpha.box = 0.3,
  labels = TRUE,
  labelsize = 3.88,
  labelround = 1,
  position = "identity",
  show.legend = NA,
  inherit.aes = TRUE,
  ...
)

GeomRocci

Arguments

mapping

Set of aesthetic mappings created by aes(). If specified and inherit.aes = TRUE (the default), it is combined with the default mapping at the top level of the plot. You must supply mapping if there is no plot mapping.

data

The data to be displayed in this layer. There are three options:

If NULL, the default, the data is inherited from the plot data as specified in the call to ggplot().

A data.frame, or other object, will override the plot data. All objects will be fortified to produce a data frame. See fortify() for which variables will be created.

A function will be called with a single argument, the plot data. The return value must be a data.frame, and will be used as the layer data. A function can be created from a formula (e.g. ~ head(.x, 10)).

stat

Use to override the default connection between geom_rocci and stat_rocci.

ci.at

Vector of values in the range of the biomarker where confidence regions will be displayed

sig.level

Significance level for the confidence regions

na.rm

If FALSE, the default, missing values are removed with a warning. If TRUE, missing values are silently removed.

alpha.box

Alpha level for the confidence regions

labels

If TRUE, adds text labels for the cutoffs where the confidence regions are displayed

labelsize

Size of cutoff text labels

labelround

Integer, number of significant digits to round cutoff labels

position

Position adjustment, either as a string naming the adjustment (e.g. "jitter" to use position_jitter), or the result of a call to a position adjustment function. Use the latter if you need to change the settings of the adjustment.

show.legend

logical. Should this layer be included in the legends? NA, the default, includes if any aesthetics are mapped. FALSE never includes, and TRUE always includes. It can also be a named logical vector to finely select the aesthetics to display.

inherit.aes

If FALSE, overrides the default aesthetics, rather than combining with them. This is most useful for helper functions that define both data and aesthetics and shouldn't inherit behaviour from the default plot specification, e.g. borders().

...

Other arguments passed on to layer(). These are often aesthetics, used to set an aesthetic to a fixed value, like colour = "red" or size = 3. They may also be parameters to the paired geom/stat.

Format

An object of class GeomRocci (inherits from Geom, ggproto, gg) of length 6.

Aesthetics

geom_rocci understands the following aesthetics (required aesthetics are in bold). stat_rocci automatically maps the estimates to the required aesthetics:

See Also

See geom_roc for the empirical ROC curve, style_roc for adding guidelines and labels, and direct_label for adding direct labels to the curves. Also export_interactive_roc for creating interactive ROC curve plots for use in a web browser.

Examples


D.ex <- rbinom(50, 1, .5)
rocdata <- data.frame(D = c(D.ex, D.ex), 
                   M = c(rnorm(50, mean = D.ex, sd = .4), rnorm(50, mean = D.ex, sd = 1)), 
                   Z = c(rep("A", 50), rep("B", 50)))

ggplot(rocdata, aes(m = M, d = D)) + geom_roc() + geom_rocci()
ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc() + geom_rocci()
ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc() + geom_rocci(sig.level = .01)
ggplot(rocdata, aes(m = M, d = D)) + geom_roc(n.cuts = 0) + 
geom_rocci(ci.at = quantile(rocdata$M, c(.1, .25, .5, .75, .9)))
ggplot(rocdata, aes(m = M, d = D, color = Z)) + geom_roc() + geom_rocci(linetype = 1)

Reads included JavaScript functions and returns them as a string for pasting into a webpage

Description

Reads included JavaScript functions and returns them as a string for pasting into a webpage

Usage

getD3()

Plot an ROC curve

Description

Deprecated, use geom_roc instead

Usage

ggroc(
  rocdata,
  fpf_string = "FPF",
  tpf_string = "TPF",
  c_string = "c",
  ci = FALSE,
  label = NULL,
  label.adj.x = 0,
  label.adj.y = 0,
  label.angle = 45,
  plotmath = FALSE,
  xlabel = "False positive fraction",
  ylabel = "True positive fraction"
)

Arguments

rocdata

Data frame containing true and false positive fractions, and cutoff values

fpf_string

Column name identifying false positive fraction column

tpf_string

Column name identifying true positive fraction column

c_string

Column name identifying cutoff values

ci

Logical, not supported

label

Not supported

label.adj.x

Not supported

label.adj.y

Not supported

label.angle

Not supported

plotmath

Not supported

xlabel

Defaults to "False positive fraction"

ylabel

Defaults to "True positive fraction"

Value

A ggplot object

Transform biomarkers stored as wide to long

Description

Multiple biomarkers measured on the same subjects are often stored as multiple columns in a data frame. This is a convenience function that transforms the data into long format, suitable for use with ggplot and geom_roc

Usage

melt_roc(data, d, m, names = NULL)

Arguments

data

Data frame containing disease status and biomarkers stored in columns

d

Column containing binary disease status. Can be a column name or index

m

Vector of column names or indices identifying biomarkers

names

Optional vector of names to assign to the biomarkers. If NULL, names will be taken from the column names

Value

A data frame in long format with three columns: D = binary disease status, M = biomarker value, and name = biomarker name

Examples

D.ex <- rbinom(50, 1, .5)
widedata <- data.frame(D = D.ex, M1 = rnorm(50, mean = D.ex, sd = 1), 
   M2 = rnorm(50, mean = D.ex, sd = .5))
longdata <- melt_roc(widedata, "D", c("M1", "M2"))
ggplot(longdata, aes(d = D, m = M, color = name)) + geom_roc()

Plot multiple ROC curves

Description

Given a list of results computed by calculate_roc, plot the curve using ggplot with sensible defaults. Pass the resulting object and data to export_interactive_roc, plot_interactive_roc, or plot_journal_roc.

Usage

multi_ggroc(
  datalist,
  fpf_string = rep("FPF", length(datalist)),
  tpf_string = rep("TPF", length(datalist)),
  c_string = rep("c", length(datalist)),
  label = NULL,
  legend = TRUE,
  label.adj.x = rep(0, length(datalist)),
  label.adj.y = rep(0, length(datalist)),
  label.angle = rep(45, length(datalist)),
  plotmath = FALSE,
  xlabel = "False positive fraction",
  ylabel = "True positive fraction"
)

Arguments

datalist

List of data frames each containing true and false positive fractions and cutoffs

fpf_string

Column names identifying false positive fraction

tpf_string

Column names identifying true positive fraction

c_string

Column names identifying cutoff values

label

Not supported.

legend

If true, draws legend

label.adj.x

Not supported.

label.adj.y

Not supported.

label.angle

Not supported.

plotmath

Logical. Not supported.

xlabel

Defaults to "False positive fraction"

ylabel

Defaults to "True positive fraction"

Value

A ggplot object

Generate a standalone html document displaying an interactive ROC curve

Description

Generate a standalone html document displaying an interactive ROC curve

Usage

plot_interactive_roc(ggroc, file = NULL, ...)

Arguments

ggroc

An object as returned by ggroc or multi_ggroc. It can be modified with annotations, themes, etc.

file

A path to save the result to. If NULL, will save to a temporary directory

...

arguments passed to export_interactive_roc

Value

NULL opens an interactive document in Rstudio or the default web browser

Plot an ROC curve for use in print

Description

Deprecated, use style_roc instead

Usage

plot_journal_roc(
  ggroc_p,
  font.size = 3,
  n.cuts = 20,
  ci.at = NULL,
  opacity = 0.3,
  lty = NULL,
  color = NULL,
  lwd = NULL,
  legend = FALSE
)

Arguments

ggroc_p

An object as returned by ggroc or multi_ggroc. It can be modified with annotations, themes, etc.

font.size

Not supported

n.cuts

Not supported

ci.at

Not supported

opacity

Not supported

lty

Not supported

color

Not supported

lwd

Not supported

legend

Not supported

Value

A ggplot object

Key for ROC geom

Description

Key for ROC geom

Usage

roc_key(data, params, size)

Arguments

data

Data created by stat

params

parameters

size

Size

Start the plotROC Shiny app

Description

A convenience function to easily start the shiny application. It will open in Rstudio, or in the default web browser.

Usage

shiny_plotROC()

Add guides and annotations to a ROC plot

Description

Adds a diagonal guideline, minor grid lines, and optionally direct labels to ggplot objects containing a geom_roc layer.

Usage

style_roc(
  major.breaks = c(0, 0.1, 0.25, 0.5, 0.75, 0.9, 1),
  minor.breaks = c(seq(0, 0.1, by = 0.01), seq(0.9, 1, by = 0.01)),
  guide = TRUE,
  xlab = "False positive fraction",
  ylab = "True positive fraction",
  theme = theme_bw
)

Arguments

major.breaks

vector of breakpoints for the major gridlines and axes

minor.breaks

vector of breakpoints for the minor gridlines and axes

guide

logical, if TRUE draws diagonal guideline

xlab

X-axis label

ylab

Y-axis label

theme

Theme function compatible with ggplot2

Examples

D.ex <- rbinom(50, 1, .5)
fakedata <- data.frame(M1 = rnorm(50, mean = D.ex), 
   D = D.ex)
ggplot(fakedata, aes(m = M1, d = D)) + geom_roc() + style_roc()
ggplot(fakedata, aes(m = M1, d = D)) + geom_roc() + style_roc(xlab = "1 - Specificity")
ggplot(fakedata, aes(m = M1, d = D)) + geom_roc() + style_roc(theme = theme_grey)

Check that D is suitable for using as binary disease status

Description

Checks for two classes and gives a warning message indicating which level is assumed to be 0/1. Throws an error if more than two levels appear in D.

Usage

verify_d(D)

Arguments

D

Vector that will be checked for 2-class labels

Value

A vector the same length as D that takes values 0, indicating no disease or 1 indicating disease.

Examples


verify_d(c(1, 0, 1))
## Not run: 
verify_d(c(TRUE, FALSE, TRUE)) #warning
verify_d(c("Dead", "Alive", "Dead")) #warning
verify_d(c("Disease", "Healthy", "Missing")) #error

## End(Not run)