R/bayesdca.h.R
In sbalci/ClinicoPathJamoviModule: Analysis for Clinicopathological Research
Defines functions
bayesdca
Documented in
bayesdca
# This file is automatically generated, you probably don't want to edit this
bayesdcaOptions <- if (requireNamespace("jmvcore", quietly=TRUE)) R6::R6Class(
"bayesdcaOptions",
inherit = jmvcore::Options,
public = list(
initialize = function(
outcomes = NULL,
outcomePos = NULL,
predictors = NULL,
thresholdMin = 0.01,
thresholdMax = 0.5,
thresholdPoints = 50,
useExternalPrevalence = FALSE,
externalCases = 100,
externalTotal = 500,
bayesianAnalysis = TRUE,
priorStrength = 2,
bootstrapCI = TRUE,
bootstrapReps = 2000,
calculateEVPI = FALSE,
nDraws = 2000,
directionIndicator = ">=", ...) {
super$initialize(
package="ClinicoPath",
name="bayesdca",
requiresData=TRUE,
...)
private$..outcomes <- jmvcore::OptionVariable$new(
"outcomes",
outcomes,
suggested=list(
"nominal"),
permitted=list(
"factor"))
private$..outcomePos <- jmvcore::OptionLevel$new(
"outcomePos",
outcomePos,
variable="(outcomes)")
private$..predictors <- jmvcore::OptionVariables$new(
"predictors",
predictors,
suggested=list(
"continuous",
"nominal"),
permitted=list(
"numeric",
"factor"))
private$..thresholdMin <- jmvcore::OptionNumber$new(
"thresholdMin",
thresholdMin,
default=0.01,
min=0.001,
max=0.5)
private$..thresholdMax <- jmvcore::OptionNumber$new(
"thresholdMax",
thresholdMax,
default=0.5,
min=0.01,
max=0.99)
private$..thresholdPoints <- jmvcore::OptionInteger$new(
"thresholdPoints",
thresholdPoints,
default=50,
min=10,
max=100)
private$..useExternalPrevalence <- jmvcore::OptionBool$new(
"useExternalPrevalence",
useExternalPrevalence,
default=FALSE)
private$..externalCases <- jmvcore::OptionInteger$new(
"externalCases",
externalCases,
default=100,
min=1)
private$..externalTotal <- jmvcore::OptionInteger$new(
"externalTotal",
externalTotal,
default=500,
min=2)
private$..bayesianAnalysis <- jmvcore::OptionBool$new(
"bayesianAnalysis",
bayesianAnalysis,
default=TRUE)
private$..priorStrength <- jmvcore::OptionNumber$new(
"priorStrength",
priorStrength,
default=2,
min=0.1,
max=10)
private$..bootstrapCI <- jmvcore::OptionBool$new(
"bootstrapCI",
bootstrapCI,
default=TRUE)
private$..bootstrapReps <- jmvcore::OptionInteger$new(
"bootstrapReps",
bootstrapReps,
default=2000,
min=500,
max=10000)
private$..calculateEVPI <- jmvcore::OptionBool$new(
"calculateEVPI",
calculateEVPI,
default=FALSE)
private$..nDraws <- jmvcore::OptionInteger$new(
"nDraws",
nDraws,
default=2000,
min=500,
max=10000)
private$..directionIndicator <- jmvcore::OptionList$new(
"directionIndicator",
directionIndicator,
options=list(
">=",
"<="),
default=">=")
self$.addOption(private$..outcomes)
self$.addOption(private$..outcomePos)
self$.addOption(private$..predictors)
self$.addOption(private$..thresholdMin)
self$.addOption(private$..thresholdMax)
self$.addOption(private$..thresholdPoints)
self$.addOption(private$..useExternalPrevalence)
self$.addOption(private$..externalCases)
self$.addOption(private$..externalTotal)
self$.addOption(private$..bayesianAnalysis)
self$.addOption(private$..priorStrength)
self$.addOption(private$..bootstrapCI)
self$.addOption(private$..bootstrapReps)
self$.addOption(private$..calculateEVPI)
self$.addOption(private$..nDraws)
self$.addOption(private$..directionIndicator)
}),
active = list(
outcomes = function() private$..outcomes$value,
outcomePos = function() private$..outcomePos$value,
predictors = function() private$..predictors$value,
thresholdMin = function() private$..thresholdMin$value,
thresholdMax = function() private$..thresholdMax$value,
thresholdPoints = function() private$..thresholdPoints$value,
useExternalPrevalence = function() private$..useExternalPrevalence$value,
externalCases = function() private$..externalCases$value,
externalTotal = function() private$..externalTotal$value,
bayesianAnalysis = function() private$..bayesianAnalysis$value,
priorStrength = function() private$..priorStrength$value,
bootstrapCI = function() private$..bootstrapCI$value,
bootstrapReps = function() private$..bootstrapReps$value,
calculateEVPI = function() private$..calculateEVPI$value,
nDraws = function() private$..nDraws$value,
directionIndicator = function() private$..directionIndicator$value),
private = list(
..outcomes = NA,
..outcomePos = NA,
..predictors = NA,
..thresholdMin = NA,
..thresholdMax = NA,
..thresholdPoints = NA,
..useExternalPrevalence = NA,
..externalCases = NA,
..externalTotal = NA,
..bayesianAnalysis = NA,
..priorStrength = NA,
..bootstrapCI = NA,
..bootstrapReps = NA,
..calculateEVPI = NA,
..nDraws = NA,
..directionIndicator = NA)
)
bayesdcaResults <- if (requireNamespace("jmvcore", quietly=TRUE)) R6::R6Class(
"bayesdcaResults",
inherit = jmvcore::Group,
active = list(
instructions = function() private$.items[["instructions"]],
summary = function() private$.items[["summary"]],
netBenefitTable = function() private$.items[["netBenefitTable"]],
modelResults = function() private$.items[["modelResults"]],
comparisonTable = function() private$.items[["comparisonTable"]],
evpiTable = function() private$.items[["evpiTable"]],
mainPlot = function() private$.items[["mainPlot"]],
deltaPlot = function() private$.items[["deltaPlot"]],
probPlot = function() private$.items[["probPlot"]],
evpiPlot = function() private$.items[["evpiPlot"]]),
private = list(),
public=list(
initialize=function(options) {
super$initialize(
options=options,
name="",
title="Bayesian Decision Curve Analysis")
self$add(jmvcore::Html$new(
options=options,
name="instructions",
title="Instructions",
visible=TRUE))
self$add(jmvcore::Html$new(
options=options,
name="summary",
title="Analysis Summary",
visible=TRUE))
self$add(jmvcore::Table$new(
options=options,
name="netBenefitTable",
title="Net Benefit Results",
visible=TRUE,
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="treatAll",
`title`="Treat All",
`type`="number",
`format`="zto"),
list(
`name`="treatNone",
`title`="Treat None",
`type`="number",
`format`="zto"))))
self$add(jmvcore::Array$new(
options=options,
name="modelResults",
title="Model/Test Results",
visible=TRUE,
template=jmvcore::Table$new(
options=options,
title="$key",
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="netBenefit",
`title`="Net Benefit",
`type`="number",
`format`="zto"),
list(
`name`="lowerCI",
`title`="Lower CI",
`type`="number",
`format`="zto",
`visible`="(bayesianAnalysis || bootstrapCI)"),
list(
`name`="upperCI",
`title`="Upper CI",
`type`="number",
`format`="zto",
`visible`="(bayesianAnalysis || bootstrapCI)"),
list(
`name`="sensitivity",
`title`="Sensitivity",
`type`="number",
`format`="pc"),
list(
`name`="specificity",
`title`="Specificity",
`type`="number",
`format`="pc")))))
self$add(jmvcore::Table$new(
options=options,
name="comparisonTable",
title="Strategy Comparison",
visible=TRUE,
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="bestStrategy",
`title`="Best Strategy",
`type`="text"),
list(
`name`="diffFromNext",
`title`="Difference from Next Best",
`type`="number",
`format`="zto"),
list(
`name`="probBest",
`title`="Probability Best",
`type`="number",
`format`="pc",
`visible`="(bayesianAnalysis)"))))
self$add(jmvcore::Table$new(
options=options,
name="evpiTable",
title="Expected Value of Perfect Information",
visible="(calculateEVPI)",
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="evpi",
`title`="EVPI",
`type`="number",
`format`="zto"))))
self$add(jmvcore::Image$new(
options=options,
name="mainPlot",
title="Decision Curves",
width=600,
height=450,
renderFun=".plotDCA",
visible=TRUE))
self$add(jmvcore::Image$new(
options=options,
name="deltaPlot",
title="Net Benefit Differences",
width=600,
height=450,
renderFun=".plotDeltaNB",
visible=TRUE))
self$add(jmvcore::Image$new(
options=options,
name="probPlot",
title="Probability of Superiority",
width=600,
height=450,
renderFun=".plotProbability",
visible="(bayesianAnalysis)"))
self$add(jmvcore::Image$new(
options=options,
name="evpiPlot",
title="Expected Value of Perfect Information",
width=600,
height=450,
renderFun=".plotEVPI",
visible="(calculateEVPI)"))}))
bayesdcaBase <- if (requireNamespace("jmvcore", quietly=TRUE)) R6::R6Class(
"bayesdcaBase",
inherit = jmvcore::Analysis,
public = list(
initialize = function(options, data=NULL, datasetId="", analysisId="", revision=0) {
super$initialize(
package = "ClinicoPath",
name = "bayesdca",
version = c(0,0,3),
options = options,
results = bayesdcaResults$new(options=options),
data = data,
datasetId = datasetId,
analysisId = analysisId,
revision = revision,
pause = NULL,
completeWhenFilled = FALSE,
requiresMissings = FALSE,
weightsSupport = 'auto')
}))
#' Bayesian Decision Curve Analysis
#'
#'
#' @param data The data as a data frame.
#' @param outcomes Binary outcome variable (0/1) representing the true disease
#' status or event.
#' @param outcomePos Specifies which level of the outcome variable should be
#' treated as the positive class.
#' @param predictors Variables containing either probability predictions from
#' models or binary results (0/1) from diagnostic tests.
#' @param thresholdMin Minimum decision threshold for the analysis.
#' @param thresholdMax Maximum decision threshold for the analysis.
#' @param thresholdPoints Number of threshold points to evaluate.
#' @param useExternalPrevalence Use external prevalence data instead of sample
#' prevalence.
#' @param externalCases Number of cases in external prevalence data.
#' @param externalTotal Total sample size in external prevalence data.
#' @param bayesianAnalysis Perform Bayesian analysis with uncertainty
#' quantification.
#' @param priorStrength Strength of prior (effective sample size).
#' @param bootstrapCI Calculate bootstrap confidence intervals for
#' non-Bayesian analysis.
#' @param bootstrapReps Number of bootstrap replications for confidence
#' intervals.
#' @param calculateEVPI Calculate Expected Value of Perfect Information.
#' @param nDraws Number of posterior draws for Bayesian analysis.
#' @param directionIndicator Direction of classification relative to the
#' cutpoint. Use '>=' when higher values predict positive outcomes.
#' @return A results object containing:
#' \tabular{llllll}{
#' \code{results$instructions} \tab \tab \tab \tab \tab a html \cr
#' \code{results$summary} \tab \tab \tab \tab \tab a html \cr
#' \code{results$netBenefitTable} \tab \tab \tab \tab \tab a table \cr
#' \code{results$modelResults} \tab \tab \tab \tab \tab an array of tables \cr
#' \code{results$comparisonTable} \tab \tab \tab \tab \tab a table \cr
#' \code{results$evpiTable} \tab \tab \tab \tab \tab a table \cr
#' \code{results$mainPlot} \tab \tab \tab \tab \tab an image \cr
#' \code{results$deltaPlot} \tab \tab \tab \tab \tab an image \cr
#' \code{results$probPlot} \tab \tab \tab \tab \tab an image \cr
#' \code{results$evpiPlot} \tab \tab \tab \tab \tab an image \cr
#' }
#'
#' Tables can be converted to data frames with \code{asDF} or \code{\link{as.data.frame}}. For example:
#'
#' \code{results$netBenefitTable$asDF}
#'
#' \code{as.data.frame(results$netBenefitTable)}
#'
#' @export
bayesdca <- function(
data,
outcomes,
outcomePos,
predictors,
thresholdMin = 0.01,
thresholdMax = 0.5,
thresholdPoints = 50,
useExternalPrevalence = FALSE,
externalCases = 100,
externalTotal = 500,
bayesianAnalysis = TRUE,
priorStrength = 2,
bootstrapCI = TRUE,
bootstrapReps = 2000,
calculateEVPI = FALSE,
nDraws = 2000,
directionIndicator = ">=") {
if ( ! requireNamespace("jmvcore", quietly=TRUE))
stop("bayesdca requires jmvcore to be installed (restart may be required)")
if ( ! missing(outcomes)) outcomes <- jmvcore::resolveQuo(jmvcore::enquo(outcomes))
if ( ! missing(predictors)) predictors <- jmvcore::resolveQuo(jmvcore::enquo(predictors))
if (missing(data))
data <- jmvcore::marshalData(
parent.frame(),
`if`( ! missing(outcomes), outcomes, NULL),
`if`( ! missing(predictors), predictors, NULL))
for (v in outcomes) if (v %in% names(data)) data[[v]] <- as.factor(data[[v]])
options <- bayesdcaOptions$new(
outcomes = outcomes,
outcomePos = outcomePos,
predictors = predictors,
thresholdMin = thresholdMin,
thresholdMax = thresholdMax,
thresholdPoints = thresholdPoints,
useExternalPrevalence = useExternalPrevalence,
externalCases = externalCases,
externalTotal = externalTotal,
bayesianAnalysis = bayesianAnalysis,
priorStrength = priorStrength,
bootstrapCI = bootstrapCI,
bootstrapReps = bootstrapReps,
calculateEVPI = calculateEVPI,
nDraws = nDraws,
directionIndicator = directionIndicator)
analysis <- bayesdcaClass$new(
options = options,
data = data)
analysis$run()
analysis$results
}
sbalci/ClinicoPathJamoviModule documentation built on June 13, 2025, 9:34 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions bayesdca
Documented in bayesdca
# This file is automatically generated, you probably don't want to edit this
bayesdcaOptions <- if (requireNamespace("jmvcore", quietly=TRUE)) R6::R6Class(
"bayesdcaOptions",
inherit = jmvcore::Options,
public = list(
initialize = function(
outcomes = NULL,
outcomePos = NULL,
predictors = NULL,
thresholdMin = 0.01,
thresholdMax = 0.5,
thresholdPoints = 50,
useExternalPrevalence = FALSE,
externalCases = 100,
externalTotal = 500,
bayesianAnalysis = TRUE,
priorStrength = 2,
bootstrapCI = TRUE,
bootstrapReps = 2000,
calculateEVPI = FALSE,
nDraws = 2000,
directionIndicator = ">=", ...) {
super$initialize(
package="ClinicoPath",
name="bayesdca",
requiresData=TRUE,
...)
private$..outcomes <- jmvcore::OptionVariable$new(
"outcomes",
outcomes,
suggested=list(
"nominal"),
permitted=list(
"factor"))
private$..outcomePos <- jmvcore::OptionLevel$new(
"outcomePos",
outcomePos,
variable="(outcomes)")
private$..predictors <- jmvcore::OptionVariables$new(
"predictors",
predictors,
suggested=list(
"continuous",
"nominal"),
permitted=list(
"numeric",
"factor"))
private$..thresholdMin <- jmvcore::OptionNumber$new(
"thresholdMin",
thresholdMin,
default=0.01,
min=0.001,
max=0.5)
private$..thresholdMax <- jmvcore::OptionNumber$new(
"thresholdMax",
thresholdMax,
default=0.5,
min=0.01,
max=0.99)
private$..thresholdPoints <- jmvcore::OptionInteger$new(
"thresholdPoints",
thresholdPoints,
default=50,
min=10,
max=100)
private$..useExternalPrevalence <- jmvcore::OptionBool$new(
"useExternalPrevalence",
useExternalPrevalence,
default=FALSE)
private$..externalCases <- jmvcore::OptionInteger$new(
"externalCases",
externalCases,
default=100,
min=1)
private$..externalTotal <- jmvcore::OptionInteger$new(
"externalTotal",
externalTotal,
default=500,
min=2)
private$..bayesianAnalysis <- jmvcore::OptionBool$new(
"bayesianAnalysis",
bayesianAnalysis,
default=TRUE)
private$..priorStrength <- jmvcore::OptionNumber$new(
"priorStrength",
priorStrength,
default=2,
min=0.1,
max=10)
private$..bootstrapCI <- jmvcore::OptionBool$new(
"bootstrapCI",
bootstrapCI,
default=TRUE)
private$..bootstrapReps <- jmvcore::OptionInteger$new(
"bootstrapReps",
bootstrapReps,
default=2000,
min=500,
max=10000)
private$..calculateEVPI <- jmvcore::OptionBool$new(
"calculateEVPI",
calculateEVPI,
default=FALSE)
private$..nDraws <- jmvcore::OptionInteger$new(
"nDraws",
nDraws,
default=2000,
min=500,
max=10000)
private$..directionIndicator <- jmvcore::OptionList$new(
"directionIndicator",
directionIndicator,
options=list(
">=",
"<="),
default=">=")
self$.addOption(private$..outcomes)
self$.addOption(private$..outcomePos)
self$.addOption(private$..predictors)
self$.addOption(private$..thresholdMin)
self$.addOption(private$..thresholdMax)
self$.addOption(private$..thresholdPoints)
self$.addOption(private$..useExternalPrevalence)
self$.addOption(private$..externalCases)
self$.addOption(private$..externalTotal)
self$.addOption(private$..bayesianAnalysis)
self$.addOption(private$..priorStrength)
self$.addOption(private$..bootstrapCI)
self$.addOption(private$..bootstrapReps)
self$.addOption(private$..calculateEVPI)
self$.addOption(private$..nDraws)
self$.addOption(private$..directionIndicator)
}),
active = list(
outcomes = function() private$..outcomes$value,
outcomePos = function() private$..outcomePos$value,
predictors = function() private$..predictors$value,
thresholdMin = function() private$..thresholdMin$value,
thresholdMax = function() private$..thresholdMax$value,
thresholdPoints = function() private$..thresholdPoints$value,
useExternalPrevalence = function() private$..useExternalPrevalence$value,
externalCases = function() private$..externalCases$value,
externalTotal = function() private$..externalTotal$value,
bayesianAnalysis = function() private$..bayesianAnalysis$value,
priorStrength = function() private$..priorStrength$value,
bootstrapCI = function() private$..bootstrapCI$value,
bootstrapReps = function() private$..bootstrapReps$value,
calculateEVPI = function() private$..calculateEVPI$value,
nDraws = function() private$..nDraws$value,
directionIndicator = function() private$..directionIndicator$value),
private = list(
..outcomes = NA,
..outcomePos = NA,
..predictors = NA,
..thresholdMin = NA,
..thresholdMax = NA,
..thresholdPoints = NA,
..useExternalPrevalence = NA,
..externalCases = NA,
..externalTotal = NA,
..bayesianAnalysis = NA,
..priorStrength = NA,
..bootstrapCI = NA,
..bootstrapReps = NA,
..calculateEVPI = NA,
..nDraws = NA,
..directionIndicator = NA)
)
bayesdcaResults <- if (requireNamespace("jmvcore", quietly=TRUE)) R6::R6Class(
"bayesdcaResults",
inherit = jmvcore::Group,
active = list(
instructions = function() private$.items[["instructions"]],
summary = function() private$.items[["summary"]],
netBenefitTable = function() private$.items[["netBenefitTable"]],
modelResults = function() private$.items[["modelResults"]],
comparisonTable = function() private$.items[["comparisonTable"]],
evpiTable = function() private$.items[["evpiTable"]],
mainPlot = function() private$.items[["mainPlot"]],
deltaPlot = function() private$.items[["deltaPlot"]],
probPlot = function() private$.items[["probPlot"]],
evpiPlot = function() private$.items[["evpiPlot"]]),
private = list(),
public=list(
initialize=function(options) {
super$initialize(
options=options,
name="",
title="Bayesian Decision Curve Analysis")
self$add(jmvcore::Html$new(
options=options,
name="instructions",
title="Instructions",
visible=TRUE))
self$add(jmvcore::Html$new(
options=options,
name="summary",
title="Analysis Summary",
visible=TRUE))
self$add(jmvcore::Table$new(
options=options,
name="netBenefitTable",
title="Net Benefit Results",
visible=TRUE,
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="treatAll",
`title`="Treat All",
`type`="number",
`format`="zto"),
list(
`name`="treatNone",
`title`="Treat None",
`type`="number",
`format`="zto"))))
self$add(jmvcore::Array$new(
options=options,
name="modelResults",
title="Model/Test Results",
visible=TRUE,
template=jmvcore::Table$new(
options=options,
title="$key",
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="netBenefit",
`title`="Net Benefit",
`type`="number",
`format`="zto"),
list(
`name`="lowerCI",
`title`="Lower CI",
`type`="number",
`format`="zto",
`visible`="(bayesianAnalysis || bootstrapCI)"),
list(
`name`="upperCI",
`title`="Upper CI",
`type`="number",
`format`="zto",
`visible`="(bayesianAnalysis || bootstrapCI)"),
list(
`name`="sensitivity",
`title`="Sensitivity",
`type`="number",
`format`="pc"),
list(
`name`="specificity",
`title`="Specificity",
`type`="number",
`format`="pc")))))
self$add(jmvcore::Table$new(
options=options,
name="comparisonTable",
title="Strategy Comparison",
visible=TRUE,
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="bestStrategy",
`title`="Best Strategy",
`type`="text"),
list(
`name`="diffFromNext",
`title`="Difference from Next Best",
`type`="number",
`format`="zto"),
list(
`name`="probBest",
`title`="Probability Best",
`type`="number",
`format`="pc",
`visible`="(bayesianAnalysis)"))))
self$add(jmvcore::Table$new(
options=options,
name="evpiTable",
title="Expected Value of Perfect Information",
visible="(calculateEVPI)",
columns=list(
list(
`name`="threshold",
`title`="Threshold",
`type`="number",
`format`="pc"),
list(
`name`="evpi",
`title`="EVPI",
`type`="number",
`format`="zto"))))
self$add(jmvcore::Image$new(
options=options,
name="mainPlot",
title="Decision Curves",
width=600,
height=450,
renderFun=".plotDCA",
visible=TRUE))
self$add(jmvcore::Image$new(
options=options,
name="deltaPlot",
title="Net Benefit Differences",
width=600,
height=450,
renderFun=".plotDeltaNB",
visible=TRUE))
self$add(jmvcore::Image$new(
options=options,
name="probPlot",
title="Probability of Superiority",
width=600,
height=450,
renderFun=".plotProbability",
visible="(bayesianAnalysis)"))
self$add(jmvcore::Image$new(
options=options,
name="evpiPlot",
title="Expected Value of Perfect Information",
width=600,
height=450,
renderFun=".plotEVPI",
visible="(calculateEVPI)"))}))
bayesdcaBase <- if (requireNamespace("jmvcore", quietly=TRUE)) R6::R6Class(
"bayesdcaBase",
inherit = jmvcore::Analysis,
public = list(
initialize = function(options, data=NULL, datasetId="", analysisId="", revision=0) {
super$initialize(
package = "ClinicoPath",
name = "bayesdca",
version = c(0,0,3),
options = options,
results = bayesdcaResults$new(options=options),
data = data,
datasetId = datasetId,
analysisId = analysisId,
revision = revision,
pause = NULL,
completeWhenFilled = FALSE,
requiresMissings = FALSE,
weightsSupport = 'auto')
}))
#' Bayesian Decision Curve Analysis
#'
#'
#' @param data The data as a data frame.
#' @param outcomes Binary outcome variable (0/1) representing the true disease
#' status or event.
#' @param outcomePos Specifies which level of the outcome variable should be
#' treated as the positive class.
#' @param predictors Variables containing either probability predictions from
#' models or binary results (0/1) from diagnostic tests.
#' @param thresholdMin Minimum decision threshold for the analysis.
#' @param thresholdMax Maximum decision threshold for the analysis.
#' @param thresholdPoints Number of threshold points to evaluate.
#' @param useExternalPrevalence Use external prevalence data instead of sample
#' prevalence.
#' @param externalCases Number of cases in external prevalence data.
#' @param externalTotal Total sample size in external prevalence data.
#' @param bayesianAnalysis Perform Bayesian analysis with uncertainty
#' quantification.
#' @param priorStrength Strength of prior (effective sample size).
#' @param bootstrapCI Calculate bootstrap confidence intervals for
#' non-Bayesian analysis.
#' @param bootstrapReps Number of bootstrap replications for confidence
#' intervals.
#' @param calculateEVPI Calculate Expected Value of Perfect Information.
#' @param nDraws Number of posterior draws for Bayesian analysis.
#' @param directionIndicator Direction of classification relative to the
#' cutpoint. Use '>=' when higher values predict positive outcomes.
#' @return A results object containing:
#' \tabular{llllll}{
#' \code{results$instructions} \tab \tab \tab \tab \tab a html \cr
#' \code{results$summary} \tab \tab \tab \tab \tab a html \cr
#' \code{results$netBenefitTable} \tab \tab \tab \tab \tab a table \cr
#' \code{results$modelResults} \tab \tab \tab \tab \tab an array of tables \cr
#' \code{results$comparisonTable} \tab \tab \tab \tab \tab a table \cr
#' \code{results$evpiTable} \tab \tab \tab \tab \tab a table \cr
#' \code{results$mainPlot} \tab \tab \tab \tab \tab an image \cr
#' \code{results$deltaPlot} \tab \tab \tab \tab \tab an image \cr
#' \code{results$probPlot} \tab \tab \tab \tab \tab an image \cr
#' \code{results$evpiPlot} \tab \tab \tab \tab \tab an image \cr
#' }
#'
#' Tables can be converted to data frames with \code{asDF} or \code{\link{as.data.frame}}. For example:
#'
#' \code{results$netBenefitTable$asDF}
#'
#' \code{as.data.frame(results$netBenefitTable)}
#'
#' @export
bayesdca <- function(
data,
outcomes,
outcomePos,
predictors,
thresholdMin = 0.01,
thresholdMax = 0.5,
thresholdPoints = 50,
useExternalPrevalence = FALSE,
externalCases = 100,
externalTotal = 500,
bayesianAnalysis = TRUE,
priorStrength = 2,
bootstrapCI = TRUE,
bootstrapReps = 2000,
calculateEVPI = FALSE,
nDraws = 2000,
directionIndicator = ">=") {
if ( ! requireNamespace("jmvcore", quietly=TRUE))
stop("bayesdca requires jmvcore to be installed (restart may be required)")
if ( ! missing(outcomes)) outcomes <- jmvcore::resolveQuo(jmvcore::enquo(outcomes))
if ( ! missing(predictors)) predictors <- jmvcore::resolveQuo(jmvcore::enquo(predictors))
if (missing(data))
data <- jmvcore::marshalData(
parent.frame(),
`if`( ! missing(outcomes), outcomes, NULL),
`if`( ! missing(predictors), predictors, NULL))
for (v in outcomes) if (v %in% names(data)) data[[v]] <- as.factor(data[[v]])
options <- bayesdcaOptions$new(
outcomes = outcomes,
outcomePos = outcomePos,
predictors = predictors,
thresholdMin = thresholdMin,
thresholdMax = thresholdMax,
thresholdPoints = thresholdPoints,
useExternalPrevalence = useExternalPrevalence,
externalCases = externalCases,
externalTotal = externalTotal,
bayesianAnalysis = bayesianAnalysis,
priorStrength = priorStrength,
bootstrapCI = bootstrapCI,
bootstrapReps = bootstrapReps,
calculateEVPI = calculateEVPI,
nDraws = nDraws,
directionIndicator = directionIndicator)
analysis <- bayesdcaClass$new(
options = options,
data = data)
analysis$run()
analysis$results
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.