Nothing
R/metadata.R
In BayesianFROC: FROC Analysis by Bayesian Approaches
Defines functions
metadata_to_DrawCurve_MRMC
metadata_to_fit_MRMC
metadata_srsc_per_image
metadata_to_fit_MRMC_casewise
Documented in
metadata_srsc_per_image
metadata_to_DrawCurve_MRMC
metadata_to_fit_MRMC
metadata_to_fit_MRMC_casewise
# title -----
#' @title Create metadata for MRMC data
#'
#'
#'@description
#' The so-called \emph{false positive fraction (FPF)}
#' and the \emph{true positive fraction (TPF)}
#' are calculated from the number of hits (True Positives: TPs)
#' and the number of false alarms (False Positives: FPs)
#'
#'
#'
#'
#'
#'
#'
#'
#'
#'@details To fit a model to data,
#' we need a hit data and false data
#' formulated by both an array and a vector.
#'
#' It also calculates the so-called
#' False Positive Fractions (FPF)
#' (resp. True Positive Fractions (TPF) )
#' which are cumulative sums of false alarms (resp.
#' hits) over number of lesions or images.
#'
#' From data of number of hits
#' and false alarms,
#' we calculate the number
#' of cumulative false positives
#' and hits per image or lesion, in other words,
#' \emph{False Positive Fraction (FPF)}
#' and \emph{True Positive Fraction (TPF)}.
#' Since there are three subscripts,
#' \emph{reader}, \emph{modality},
#' and \emph{image}, we can create array format
#' or vector format etc...
#'
#' \strong{Abbreviations}
#'
#' \emph{FPF: false positive fraction }
#'
#' \emph{TPF: true positive fraction }
#'
#' \emph{hit : True Positive = TP}
#'
#' \emph{false alarms: False Positive = FP}
#'
#'
#'
#'
#'
#'
#' The traditionaly, the so-called FPF;\emph{False Positive Fraction} and TPT:\emph{True Positive Fraction} are used.
#' Recall that our data format:
#'
#'
#'
#'
#' \emph{ A single reader and a single modality case }
#'
#'------------------------------------------------------------------------------------------------------
#' \tabular{llll}{
#' \code{NI, NL } \tab \strong{ confidence level } \tab \strong{ No. of false alarms} \tab \strong{No. of hits} \cr
#' \tab \tab (FP:False Positive) \tab (TP:True Positive) \cr
#' -----------------------\tab ----------------------- \tab ----------------------------- \tab ------------- \cr
#' \emph{definitely} present \tab 5 \tab \eqn{F_5} \tab \eqn{H_5} \cr
#' \emph{probably} present \tab 4 \tab \eqn{F_4} \tab \eqn{H_4} \cr
#' equivocal \tab 3 \tab \eqn{F_3} \tab \eqn{H_3} \cr
#' subtle \tab 2 \tab \eqn{F_2} \tab \eqn{H_2} \cr
#' \emph{very} subtle \tab 1 \tab \eqn{F_1} \tab \eqn{H_1} \cr
#' }
#'
#'---------------------------------------------------------------------------------------------------
#'
#' FPF is defined as follows;
#'
#'
#' \deqn{FPF(5):= \frac{F_5}{NI},}
#' \deqn{FPF(4):= \frac{F_4+F_5}{NI},}
#' \deqn{FPF(3):= \frac{F_3+F_4+F_5}{NI},}
#' \deqn{FPF(2):= \frac{F_2+F_3+F_4+F_5}{NI},}
#' \deqn{FPF(1):= \frac{F_1+F_2+F_3+F_4+F_5}{NI}.}
#'
#'
#' TPF is defined as follows;
#'
#'
#' \deqn{TPF(5):= \frac{H_5}{NL},}
#' \deqn{TPF(4):= \frac{H_4+H_5}{NL},}
#' \deqn{TPF(3):= \frac{H_3+H_4+H_5}{NL},}
#' \deqn{TPF(2):= \frac{H_2+H_3+H_4+H_5}{NL},}
#' \deqn{TPF(1):= \frac{H_1+H_2+H_3+H_4+H_5}{NL}.}
#'
#'
#'
#'
#'
#'
#'
#'
#'@param dataList A list, consisting of the following \R objects:\code{m,q,c,h,f,NL,C,M,Q} each of which means from the right
#'
#'\code{caseID } : A vector, indicating the case ID (image, radiograph, patient ... etc) = 1,2,... which does not include zero.
#'
#'\code{m } : A vector, indicating the modality ID = 1,2,... which does not include zero.
#'
#'\code{q } : A vector, indicating the reader ID = 1,2,... which does not include zero.
#'
#'\code{c } : A vector, indicating the confidence = 1,2,... which does not include zero.
#'
#'\code{h } : A vector, indicating the number of hits
#'
#'\code{f } : A vector, indicating the number of false alarm
#'
#'\code{NL } : A positive integer, indicating the number of lesions for all images
#'
#'\code{C } : A positive integer, indicating the highest number of confidence level
#'
#'\code{M } : A positive integer, indicating the number of modalities
#'
#'\code{Q } : A positive integer, indicating the number of readers.
#'
#'The detail of these dataset, please see the example datasets, e.g. \code{\link{dd}}.
# @return-----
#'@return A list, which includes arrays and vectors.
#'A metadata such as number of cumulative false alarms
#' and hits to create and draw the curve.
#'
#' The
#' \emph{False Positive Fraction (FPF)} and
#' \emph{True Positive Fraction (TPF)} are also calculated.
#'
#'\strong{ The components of list} \emph{ I rediscover it at 2019 Jun 18, I am not sure it is useful? 2019 Dec 8}
#' \describe{
#'\item{ \code{ harray} }{An array of hit, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ farray} }{An array of false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharray} }{An array of \strong{cumulative} hits, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarray} }{An array of \strong{cumulative} false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharrayN} }{An array of TPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarrayN} }{An array of FPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ h} }{An vector of hit, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ f } }{An vector of false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hh } }{An vector of \strong{cumulative} hits, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ff } }{An vector of \strong{cumulative} false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hhN} }{An vector of TPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffN} }{An vector of FPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'}
#'
#'
#' Revised Nov. 21
#'
#'
#'
#'@examples
#'#========================================================================================
#'# First, we prepare the data endowed with this package.
#'#========================================================================================
#'
#'
#'
#' dat <- get(data("dataList.Chakra.Web"))
#'
#'
#'#========================================================================================
#'# #Calculate FPFs and TPFs and etc.
#'#========================================================================================
#'
#'
#'
#' a <- metadata_to_fit_MRMC(dat)
#'
#'
#' #Now, we get a meta-data object named "a".
#'
#'
#'#========================================================================================
#'# Check of Definiion
#'#========================================================================================
#'
#'
#' a$hh/dat$NL == a$hhN
#'
#'# Since all of aboves are TRUE, the hhN is a TPF per NL.
#'
#'
#'
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs
#'#========================================================================================
#'#'
#'
#'
#'
#' FPF = a$ffN
#' TPF = a$hhN
#'
#' dark_theme()
#' plot(FPF,TPF)
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs via ggplot
#'#========================================================================================
#'
#' length(dat$f)==length(FPF)
#'
#' q <- dat$q
#' m <- dat$m
#' df <- data.frame(FPF,
#' TPF,
#' m,
#' q
#' )
#'
#'# ggplot2::ggplot(df, aes(x =FPF, y = TPF, colour = q, group = m)) + ggplot2::geom_point()
#'
#'# Revised 2019 Jun 18, Revised 2019 Sept 9
#'
#'
#'
#'
#'
# devtools::document();help("metadata_MRMC") # Confirm reflection
# devtools::use_data(dataList.high.ability)
#' @export metadata_to_fit_MRMC
#'@inheritParams fit_Bayesian_FROC
metadata_to_fit_MRMC_casewise<- function(dataList,ModifiedPoisson=FALSE)
{
# message("\n")
# message("* Now, we calculated the metadata, e.g., cumulative hits and false alarms etc... \n")
NL_casewise <- dataList$NL_casewise
caseIDs_deseased <- dataList$caseIDs_deseased
NI_deseased <- length( caseIDs_deseased)
caseID <- dataList$caseID
m <-dataList$m
q<-dataList$q
c<-dataList$c
h<-dataList$h
f<-dataList$f
NI<-dataList$NI
NL<-dataList$NL
C<-dataList$C
M<-dataList$M
Q<-dataList$Q
modalityID_dummy <- fastDummies::dummy_cols( data.frame( modalityID_dummy=m) )
modalityID_dummy <- modalityID_dummy[, colnames(modalityID_dummy) != "modalityID_dummy"]
modalityID_dummy <- as.matrix(modalityID_dummy)
readerID_dummy <- fastDummies::dummy_cols( data.frame( readerID_dummy=q) )
readerID_dummy <- readerID_dummy[, colnames(readerID_dummy) != "readerID_dummy"]
readerID_dummy <- as.matrix(readerID_dummy)
caseID_dummy <- fastDummies::dummy_cols( data.frame( caseID_dummy=caseID) )
caseID_dummy <- caseID_dummy[, colnames(caseID_dummy) != "caseID_dummy"]
caseID_dummy <- as.matrix(caseID_dummy)
if(ModifiedPoisson==FALSE) NX = NI;
if(ModifiedPoisson==TRUE) NX =NL;
N <- NI*M*Q*C
hh <- numeric(N) #Initialization of Cumulative Hits
ff <- numeric(N) #Initialization of Cumulative False alarm
hharray<-array(0,dim=c(NI,C,M,Q));#Cumulative sum
ffarray<-array(0,dim=c(NI,C,M,Q));#Cumulative sum
harray<-array(0,dim=c(NI,C,M,Q));#Non-Cumulative
farray<-array(0,dim=c(NI,C,M,Q)); #Non-Cumulative
# hCQ<-array(0,dim=c(C,Q));#Non-Cumulative, that is, merely format is only changed from vector to array.
# fCQ<-array(0,dim=c(C,Q)); #Non-Cumulative, that is, merely format is only changed from vector to array.
# hC<-array(0,dim=c(C));#Non-Cumulative, that is, merely format is only changed from vector to array.
# fC<-array(0,dim=c(C)); #Non-Cumulative, that is, merely format is only changed from vector to array.
for(case in 1:NI){
for(mmm in 1:M) {
for(ccc in 1:C) {
for(qqq in 1 : Q){
for(n in 1:ccc){
ff[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]<-ff[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]+f[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-n)]
hh[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]<-hh[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]+h[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-n)]
}
ffarray[case,ccc,mmm,qqq] <-ff[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
hharray[case,ccc,mmm,qqq] <-hh[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
farray[case,ccc,mmm,qqq] <- f[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
harray[case,ccc,mmm,qqq] <- h[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
}}}}
# for(cd in 1:C){
# for(qd in 1:Q){
# for(md in 1:M){
# # hCQ[cd, qd]<-hCQ[cd, qd]+harray[cd,md,qd]
# # fCQ[cd, qd]<-fCQ[cd, qd]+farray[cd,md,qd]
# }}}
# for(cd in 1:C){
# for(qd in 1:Q){
# for(md in 1:M){
# # hC[cd]<-hC[cd]+harray[cd,md,qd]
# # fC[cd]<-fC[cd]+farray[cd,md,qd]
# }}}
# S <-array(0,dim=c(M,Q));
# for(md in 1:M){
# for(qd in 1 : Q){
# S[md,qd] <- sum(harray[,md,qd])
# }}
hharrayN<-hharray/NL
ffarrayN<-ffarray/NX
hhN<-hh/NL
ffN<-ff/NX
data <- list(N=N,Q=Q, M=M,m=m ,C=C ,caseID=caseID,NI_deseased=NI_deseased,
# S=S,
NL=NL,NI=NI, ccc = C:1
,c=c,q=q,
h=h, f=f,
hh=hh, hhN=hhN,
ff=ff,ffN=ffN,
harray=harray, farray=farray,
hharray=hharray, ffarray=ffarray,
hharrayN=hharrayN, ffarrayN=ffarrayN,
NL_casewise=NL_casewise,
caseIDs_deseased=caseIDs_deseased,
# Return ----
modalityID_dummy = modalityID_dummy,
readerID_dummy = readerID_dummy,
caseID_dummy = caseID_dummy
)
return(data)
}
#' @title Create metadata for MRMC data.
#'
#'@description
#' The so-called \emph{false positive fraction (FPF)}
#' and the \emph{true positive fraction (TPF)}
#' are calculated from the number of hits (True Positives: TPs)
#' and the number of false alarms (False Positives: FPs)
#'
#'
#'
#'
#'
#'
#'
#'@details
#' From data of number of hits (True Positive: TP)
#' and false alarms (False Positive: FP), we calculate the number
#' of cumulative false positives (FPF) and cumulative hits (TPF).
#'
#' Because there are three subscripts,
#' reader, modality, and image,
#' we create array format and vector format etc...
#'
#'@param dataList A list, should include
#' \code{m,q,c,h,f,NL,C,M,Q} which means
#'
#'\code{c} should be
#'created by \code{ c <-c(rep(C:1))},
#' where \code{C} is the number of confidence levels.
#'So, you should write down your hits
#'and false alarms vector so that it
#'is compatible with this
#' automatically created \code{c} vector.
#'
#'\code{h} means the number of hits
#'
#'\code{f} means the number of false alarm
#'
#'\code{NL} means the Total number of lesions for all images
#'
#'\code{C} means the highest number of confidence level
#'
#'@return A metadata such as number
#' of cumulative false alarms and
#' hits to create and draw the curve.
#'
#'
#'
#'
#'
#'
#'
#'@examples
#' \dontrun{
#'#========================================================================================
#'# TP and FP
#'#========================================================================================
#'
#'
#' dat <- BayesianFROC::dataList.Chakra.Web
#'
#'
#'
#'#========================================================================================
#'# Calculates TPF and FPF from TP and FP
#'#========================================================================================
#'
#'
#' metadata_srsc_per_image(dat)
#'
#'
#'
#'
#'
#'
#' # Revised 2019 Nov.
#'
#'
#'}# dottest
#' @export metadata_srsc_per_image
#'@inheritParams fit_Bayesian_FROC
metadata_srsc_per_image<- function(dataList,ModifiedPoisson)
{
if ( length(dataList)>=7)return(message("srsc Only. Your data is not srsc."))# This cause error in ReadMe, so It is better to omit this line. 2019 Sept
h <- dataList$h
f <- dataList$f
NI <- dataList$NI
NL <- dataList$NL
C <- as.integer(dataList$C)
#
# if(ModifiedPoisson)NX <- NL
# if(!ModifiedPoisson)NX <- NI
#
NX <- if(ModifiedPoisson) NL else if(!ModifiedPoisson) NI
# browser()
# C <-length(h)
c <-c(rep(C:1))
if ( (length(h) >length(f))
|| (length(h) <length(f))
) { return(message("Format error:\nIn your data, true positive and false positives are not same length.\n"))
} else
if ( (sum(h, na.rm = TRUE) >NL) )
{return(message("Format error:\nIn your data, number of true positives are greater than that of lesions.\n"))
} else
N <- length(f)
# hh <- numeric(N) #cumulative fraction
# ff <- numeric(N) #cumulative fraction
# for(cd in 1:C) {
# for(n in 1:cd) {
# hh[cd]<-hh[cd]+h[n]/NL #cumulative fraction to examine the fitness of FROC
# ff[cd]<-ff[cd]+f[n]/NI #cumulative fraction to examine the fitness of FROC
# }}
# fff <- numeric(N)
# for(cd in 1:C) {
# for(n in 1:cd) {
# fff[cd]<-fff[cd]+f[n] #cumulative fraction to examine the fitness of FROC
# }}
fff <- cumsum(f)
hh <- cumsum( h)/NL
ff <- fff/NX
hitExtended <- vector( length =(C+1),mode = "integer")
falsealarmExtended <- vector( length =(C+1),mode = "integer")
for (cd in 1:C) hitExtended[cd] = h[cd];
hitExtended[C+1]=NL-sum(h, na.rm = TRUE);
for (cd in 1:C) falsealarmExtended[cd] = f[cd];#2020 Dec 7
falsealarmExtended[C+1]=NI-sum(f);#2020 Dec 7
data <- list( N=N,NL=NL,NI=NI,C=as.integer(C),c=c,
h=h,f=f,
hh=hh,ff=ff,
fff=fff,
hitExtended = hitExtended,
falsealarmExtended=falsealarmExtended
)
invisible(data)
}
# title -----
#' @title Create metadata for MRMC data
#'
#'
#'@description
#' The so-called \emph{false positive fraction (FPF)}
#' and the \emph{true positive fraction (TPF)}
#' are calculated from the number of hits (True Positives: TPs)
#' and the number of false alarms (False Positives: FPs)
#'
#'
#'
#'
#'
#'
#'
#'
#'
#'@details To fit a model to data,
#' we need a hit data and false data
#' formulated by both an array and a vector.
#'
#' It also calculates the so-called
#' False Positive Fractions (FPF)
#' (resp. True Positive Fractions (TPF) )
#' which are cumulative sums of false alarms (resp.
#' hits) over number of lesions or images.
#'
#' From data of number of hits
#' and false alarms,
#' we calculate the number
#' of cumulative false positives
#' and hits per image or lesion, in other words,
#' \emph{False Positive Fraction (FPF)}
#' and \emph{True Positive Fraction (TPF)}.
#' Since there are three subscripts,
#' \emph{reader}, \emph{modality},
#' and \emph{image}, we can create array format
#' or vector format etc...
#'
#' \strong{Abbreviations}
#'
#' \emph{FPF: false positive fraction }
#'
#' \emph{TPF: true positive fraction }
#'
#' \emph{hit : True Positive = TP}
#'
#' \emph{false alarms: False Positive = FP}
#'
#'
#'
#'
#'
#'
#' The traditionaly, the so-called FPF;\emph{False Positive Fraction} and TPT:\emph{True Positive Fraction} are used.
#' Recall that our data format:
#'
#'
#'
#'
#' \emph{ A single reader and a single modality case }
#'
#'------------------------------------------------------------------------------------------------------
#' \tabular{llll}{
#' \code{NI, NL } \tab \strong{ confidence level } \tab \strong{ No. of false alarms} \tab \strong{No. of hits} \cr
#' \tab \tab (FP:False Positive) \tab (TP:True Positive) \cr
#' -----------------------\tab ----------------------- \tab ----------------------------- \tab ------------- \cr
#' \emph{definitely} present \tab 5 \tab \eqn{F_5} \tab \eqn{H_5} \cr
#' \emph{probably} present \tab 4 \tab \eqn{F_4} \tab \eqn{H_4} \cr
#' equivocal \tab 3 \tab \eqn{F_3} \tab \eqn{H_3} \cr
#' subtle \tab 2 \tab \eqn{F_2} \tab \eqn{H_2} \cr
#' \emph{very} subtle \tab 1 \tab \eqn{F_1} \tab \eqn{H_1} \cr
#' }
#'
#'---------------------------------------------------------------------------------------------------
#'
#' FPF is defined as follows;
#'
#'
#' \deqn{FPF(5):= \frac{F_5}{NI},}
#' \deqn{FPF(4):= \frac{F_4+F_5}{NI},}
#' \deqn{FPF(3):= \frac{F_3+F_4+F_5}{NI},}
#' \deqn{FPF(2):= \frac{F_2+F_3+F_4+F_5}{NI},}
#' \deqn{FPF(1):= \frac{F_1+F_2+F_3+F_4+F_5}{NI}.}
#'
#'
#' TPF is defined as follows;
#'
#'
#' \deqn{TPF(5):= \frac{H_5}{NL},}
#' \deqn{TPF(4):= \frac{H_4+H_5}{NL},}
#' \deqn{TPF(3):= \frac{H_3+H_4+H_5}{NL},}
#' \deqn{TPF(2):= \frac{H_2+H_3+H_4+H_5}{NL},}
#' \deqn{TPF(1):= \frac{H_1+H_2+H_3+H_4+H_5}{NL}.}
#'
#'
#'
#'
#'
#'
#'
#'
#'@param dataList A list, consisting of the following \R objects:\code{m,q,c,h,f,NL,C,M,Q} each of which means from the right
#'
#'\code{m } : A vector, indicating the modality ID = 1,2,... which does not include zero.
#'
#'\code{q } : A vector, indicating the reader ID = 1,2,... which does not include zero.
#'
#'\code{c } : A vector, indicating the confidence = 1,2,... which does not include zero.
#'
#'\code{h } : A vector, indicating the number of hits
#'
#'\code{f } : A vector, indicating the number of false alarm
#'
#'\code{NL } : A positive integer, indicating the number of lesions for all images
#'
#'\code{C } : A positive integer, indicating the highest number of confidence level
#'
#'\code{M } : A positive integer, indicating the number of modalities
#'
#'\code{Q } : A positive integer, indicating the number of readers.
#'
#'The detail of these dataset, please see the example datasets, e.g. \code{\link{dd}}.
# @return-----
#'@return A list, which includes arrays and vectors.
#'A metadata such as number of cumulative false alarms
#' and hits to create and draw the curve.
#'
#' The
#' \emph{False Positive Fraction (FPF)} and
#' \emph{True Positive Fraction (TPF)} are also calculated.
#'
#'\strong{ The components of list} \emph{ I rediscover it at 2019 Jun 18, I am not sure it is useful? 2019 Dec 8}
#' \describe{
#'\item{ \code{ harray} }{An array of hit, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ farray} }{An array of false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharray} }{An array of \strong{cumulative} hits, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarray} }{An array of \strong{cumulative} false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharrayN} }{An array of TPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarrayN} }{An array of FPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ h} }{An vector of hit, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ f } }{An vector of false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hh } }{An vector of \strong{cumulative} hits, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ff } }{An vector of \strong{cumulative} false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hhN} }{An vector of TPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffN} }{An vector of FPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'}
#'
#'
#' Revised Nov. 21
#'
#'
#'
#'@examples
#'#========================================================================================
#'# First, we prepare the data endowed with this package.
#'#========================================================================================
#'
#'
#'
#' dat <- get(data("dataList.Chakra.Web"))
#'
#'
#'#========================================================================================
#'# #Calculate FPFs and TPFs and etc.
#'#========================================================================================
#'
#'
#'
#' a <- metadata_to_fit_MRMC(dat)
#'
#'
#' #Now, we get a meta-data object named "a".
#'
#'
#'#========================================================================================
#'# Check of Definiion
#'#========================================================================================
#'
#'
#' a$hh/dat$NL == a$hhN
#'
#'# Since all of aboves are TRUE, the hhN is a TPF per NL.
#'
#'
#'
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs
#'#========================================================================================
#'#'
#'
#'
#'
#' FPF = a$ffN
#' TPF = a$hhN
#'
#' dark_theme()
#' plot(FPF,TPF)
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs via ggplot
#'#========================================================================================
#'
#' length(dat$f)==length(FPF)
#'
#' q <- dat$q
#' m <- dat$m
#' df <- data.frame(FPF,
#' TPF,
#' m,
#' q
#' )
#'
#'# ggplot2::ggplot(df, aes(x =FPF, y = TPF, colour = q, group = m)) + ggplot2::geom_point()
#'
#'# Revised 2019 Jun 18, Revised 2019 Sept 9
#'
#'
#'
#'
#'
# devtools::document();help("metadata_MRMC") # Confirm reflection
# devtools::use_data(dataList.high.ability)
#' @export metadata_to_fit_MRMC
#'@inheritParams fit_Bayesian_FROC
metadata_to_fit_MRMC<- function(dataList,ModifiedPoisson=FALSE)
{
# message("\n")
# message("* Now, we calculated the metadata, e.g., cumulative hits and false alarms etc... \n")
m <-dataList$m
q<-dataList$q
c<-dataList$c
h<-dataList$h
f<-dataList$f
NI<-dataList$NI
NL<-dataList$NL
C<-dataList$C
M<-dataList$M
Q<-dataList$Q
if(ModifiedPoisson==FALSE) NX = NI;
if(ModifiedPoisson==TRUE) NX =NL;
N <-C*M*Q
#For Draw the Sample points on FROC curve. Assessment of Fit for FROC.
hh <- numeric(N) #Initialization of Cumulative Hits
ff <- numeric(N) #Initialization of Cumulative False alarm
hharray<-array(0,dim=c(C,M,Q));#Cumulative
ffarray<-array(0,dim=c(C,M,Q));#Cumulative
harray<-array(0,dim=c(C,M,Q));#Non-Cumulative
farray<-array(0,dim=c(C,M,Q)); #Non-Cumulative
hCQ<-array(0,dim=c(C,Q));#Non-Cumulative, that is, merely format is only changed from vector to array.
fCQ<-array(0,dim=c(C,Q)); #Non-Cumulative, that is, merely format is only changed from vector to array.
hC<-array(0,dim=c(C));#Non-Cumulative, that is, merely format is only changed from vector to array.
fC<-array(0,dim=c(C)); #Non-Cumulative, that is, merely format is only changed from vector to array.
for(md in 1:M) {
for(cd in 1:C) {
for(qd in 1 : Q){
for(n in 1:cd){
ff[cd+(md-1)*C*Q+(qd-1)*C]<-ff[cd+(md-1)*C*Q+(qd-1)*C]+f[n+(md-1)*C*Q+(qd-1)*C]
hh[cd+(md-1)*C*Q+(qd-1)*C]<-hh[cd+(md-1)*C*Q+(qd-1)*C]+h[n+(md-1)*C*Q+(qd-1)*C]
}
ffarray[cd,md,qd] <-ff[cd+(md-1)*C*Q+(qd-1)*C]
hharray[cd,md,qd] <-hh[cd+(md-1)*C*Q+(qd-1)*C]
farray[cd,md,qd] <- f[cd+(md-1)*C*Q+(qd-1)*C]
harray[cd,md,qd] <- h[cd+(md-1)*C*Q+(qd-1)*C]
}}}
for(cd in 1:C){
for(qd in 1:Q){
for(md in 1:M){
hCQ[cd, qd]<-hCQ[cd, qd]+harray[cd,md,qd]
fCQ[cd, qd]<-fCQ[cd, qd]+farray[cd,md,qd]
}}}
for(cd in 1:C){
for(qd in 1:Q){
for(md in 1:M){
hC[cd]<-hC[cd]+harray[cd,md,qd]
fC[cd]<-fC[cd]+farray[cd,md,qd]
}}}
S <-array(0,dim=c(M,Q));
for(md in 1:M){
for(qd in 1 : Q){
S[md,qd] <- sum(harray[,md,qd])
}}
hharrayN<-hharray/NL
ffarrayN<-ffarray/NX
hhN<-hh/NL
ffN<-ff/NX
data <- list(N=N,Q=Q, M=M,m=m ,C=C ,S=S, NL=NL,NI=NI, ccc = C:1
,c=c,q=q,
h=h, f=f,
hh=hh, hhN=hhN,
ff=ff,ffN=ffN,
harray=harray, farray=farray,
hharray=hharray, ffarray=ffarray,
hharrayN=hharrayN, ffarrayN=ffarrayN)
return(data)
}
#' @title Create metadata for MRMC data
#'@description From data of number of hits and false alarms, we calculate the number of cumulative false positives and hits.
#' Since there are three subscripts, reader, modality, and image, we create array format and vector format etc...
#'@inheritParams fit_Bayesian_FROC
#'@inheritParams DrawCurves
#'@return A metadata such as number of cumulative false alarms and hits to create and draw the curve.
#' @export metadata_to_DrawCurve_MRMC
metadata_to_DrawCurve_MRMC<- function(
StanS4class,
mesh.for.drawing.curve=5000)
{
fit <-StanS4class
data <-fit@metadata
m<-data$m ;S<-data$S; NL<-data$NL;c<-data$c;q<-data$q;
h<-data$h; f<-data$f;
hh<-data$hh; hhN<-data$hhN;
ff<-data$ff;ffN<-data$ffN;
harray<-data$harray; farray<-data$farray;
hharray<-data$hharray; ffarray<-data$ffarray;
hharrayN<-data$hharrayN; ffarrayN<-data$ffarrayN;
C<-as.integer(data$C)
M<-as.integer(data$M)
N<-as.integer(data$N)
Q<-as.integer(data$Q)
# message(crayon::silver("\n* Metadata to draw the curves are callculating ... \n"))
fit <- methods::as(StanS4class, "stanfit")
war <- fit@sim$warmup
cha <- fit@sim$chains
ite <- fit@sim$iter
#Draw the AFROC curve~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
MCMC=(ite-war)*cha #Total Samples By MCMC
ll<- stats::rchisq(mesh.for.drawing.curve, 1)
lll<- 0.99+stats::rchisq(mesh.for.drawing.curve, 1)
l<-append(ll,lll) #This name l is duplicated!! CAUTION!!
#sort ----
l<-sort(l, method = "shell", index.return = FALSE)
x<- 1-exp(-l) #AFROC-Curve
y <- array(0, dim=c(length(x), M,Q)) #
# a<-rstan::extract(fit)$a #samples of a by MCMC
# b<-rstan::extract(fit)$b #samples of b by MCMC
# EAP_a <- array(0, dim=c( M,Q)) #
# EAP_b <- array(0, dim=c( M,Q)) #
#
# for(md in 1:M){
# for(qd in 1:Q){
#
# EAP_a[md,qd] <- 0
# EAP_b[md,qd] <- 0
# s<-0
# t<-0
# for(mc in 1:MCMC){ #EAP
# s<- EAP_a[md,qd]
# EAP_a[md,qd] <- s+ a[mc,md,qd]
# t<- EAP_b[md,qd]
# EAP_b[md,qd] <- t+ b[mc,md,qd]
# }
# EAP_a[md,qd] <-EAP_a[md,qd] /MCMC #EAP of a
# EAP_b[md,qd] <-EAP_b[md,qd] /MCMC #EAP of b
# }}
EAP_a <- extract_EAP_by_array(fit,"aa")
EAP_b <- extract_EAP_by_array(fit,"bb")
EAP_AA <- extract_EAP_by_array(fit,"AA")
# message("* Process for calculating y coordinates of curves \n")#Processsssssss
plotFROCdata <- list()
plotAFROCdata <- list()
nameFROC <- vector()
nameAFROC <- vector()
#Vectorization #Vectorization #Vectorization #Vectorization
for(i in 1:length(x)){ #y-coord.of FROC and AFROC are same
y[i,,]<-1-stats::pnorm(EAP_b *stats::qnorm(exp(-l[i]))-EAP_a )#20210120
}
#Vectorization #Vectorization #Vectorization #Vectorization
s <- 0
for(qd in 1:Q){
for(md in 1:M){
s<-s+1
# message("|")#Processsssssss
plotFROCdata[[s]] <- data.frame(x.coord = l, y.coord =y[,md,qd] )
plotAFROCdata[[s]] <- data.frame(x.coord = x, y.coord =y[,md,qd] )
names( plotFROCdata[[s]] ) <- c(paste("X.coord.for.FROC.curve.for.modality.",md,".and.reader.",qd,sep = ""),paste("Y.coord.for.FROC.curve.for.modality.",md,".and.reader.",qd,sep = ""))
names( plotAFROCdata[[s]] ) <- c(paste("X.coord.for.AFROC.curve.for.modality.",md,".and.reader.",qd,sep = ""),paste("Y.coord.for.AFROC.curve.for.modality.",md,".and.reader.",qd,sep = ""))
nameFROC[s] <- paste("FROC.data.for.modality.",md,".and.reader.",qd,sep = "")
nameAFROC[s] <- paste("AFROC.data.for.modality.",md,".and.reader.",qd,sep = "")
# names(list$plotAFROCdata)
}
# message(paste("", ceiling(round(qd/Q,2)*100/2),"% \n"))#Processsssssss
}
names(plotFROCdata) <-nameFROC
names(plotAFROCdata) <-nameAFROC
#aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
#-- AUC for each modarlity and each reader
# AA<-rstan::extract(fit)$AA #samples of a by MCMC
# EAP_AA <- array(0, dim=c( M,Q)) #
#
# for(qd in 1:Q){
# for(md in 1:M){
# EAP_AA[md,qd] <- mean(AA[,md,qd])
# # EAP_AA[md,qd] <- 0
# # s<-0
# # message("|")#Processsssssss
# # for(mc in 1:MCMC){ #EAP
# # s<- EAP_AA[md,qd]
# # EAP_AA[md,qd] <- s+ AA[mc,md,qd]
# # }
# # EAP_AA[md,qd] <-EAP_AA[md,qd] /MCMC #EAP of a
# }#for md
# # message(paste("", ceiling(round(qd/Q,2)*100/2 +50),"% \n"))#Processsssssss
# }# for qd
# aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
# #--------- chi ^2 -----------Start
# p<-rstan::extract(fit)$p
# lambda<-rstan::extract(fit)$l
# EAP_p <- array(0, dim=c( C,M,Q))
# EAP_l <- array(0, dim=c( C+1))
#
# s <- array(0, dim=c( C,M,Q))
# t <- array(0, dim=c( C+1))
# for(mc in 1:MCMC){
# for(cd in 1:C){
# for(md in 1:M){
# for(qd in 1:Q){
# s[ cd,md,qd]<- EAP_p[ cd,md,qd]
#
# EAP_p[ cd,md,qd] <- s[ cd,md,qd]+ p[mc,cd,md,qd]
# }
# for(cd in 0:C){
# t[ cd]<- EAP_l[ cd]
# EAP_l[ cd] <- t[ cd]+ lambda[mc,cd]
# }
# }}}
# EAP_p<-EAP_p/MCMC
# EAP_l<-EAP_l/MCMC
#
# ss<-array(0, dim=c( M,Q))
# tt<-array(0, dim=c( M,Q))
# for(cd in 1:C){
# for(md in 1:M){
# for(qd in 1:Q){
# ss[md,qd]<-(hharray[C+1-cd,md,qd]-NL*EAP_p[cd,md,qd])^2/(NL*EAP_p[cd,md,qd])
# ## tt<-(ffarray[C+1-cd,md,qd]-NI*(EAP_l[cd ]-EAP_l[cd+1 ]))^2/(NI*(EAP_l[cd ]-EAP_l[cd+1 ]))
# tt[md,qd]<-(ffarray[C+1-cd,md,qd]-NL*(EAP_l[cd ]-EAP_l[cd+1 ]))^2/(NL*(EAP_l[cd ]-EAP_l[cd+1 ]))
#
# }}}
# chisquare <- ss+tt
# chisquare
#
invisible(list(x=x,y=y,l=l,EAP_AA=EAP_AA ,plotFROCdata=plotFROCdata,plotAFROCdata=plotAFROCdata))
}
Try the BayesianFROC package in your browser
Any scripts or data that you put into this service are public.
BayesianFROC documentation built on Jan. 23, 2022, 9:06 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 metadata_to_DrawCurve_MRMC metadata_to_fit_MRMC metadata_srsc_per_image metadata_to_fit_MRMC_casewise
Documented in metadata_srsc_per_image metadata_to_DrawCurve_MRMC metadata_to_fit_MRMC metadata_to_fit_MRMC_casewise
# title -----
#' @title Create metadata for MRMC data
#'
#'
#'@description
#' The so-called \emph{false positive fraction (FPF)}
#' and the \emph{true positive fraction (TPF)}
#' are calculated from the number of hits (True Positives: TPs)
#' and the number of false alarms (False Positives: FPs)
#'
#'
#'
#'
#'
#'
#'
#'
#'
#'@details To fit a model to data,
#' we need a hit data and false data
#' formulated by both an array and a vector.
#'
#' It also calculates the so-called
#' False Positive Fractions (FPF)
#' (resp. True Positive Fractions (TPF) )
#' which are cumulative sums of false alarms (resp.
#' hits) over number of lesions or images.
#'
#' From data of number of hits
#' and false alarms,
#' we calculate the number
#' of cumulative false positives
#' and hits per image or lesion, in other words,
#' \emph{False Positive Fraction (FPF)}
#' and \emph{True Positive Fraction (TPF)}.
#' Since there are three subscripts,
#' \emph{reader}, \emph{modality},
#' and \emph{image}, we can create array format
#' or vector format etc...
#'
#' \strong{Abbreviations}
#'
#' \emph{FPF: false positive fraction }
#'
#' \emph{TPF: true positive fraction }
#'
#' \emph{hit : True Positive = TP}
#'
#' \emph{false alarms: False Positive = FP}
#'
#'
#'
#'
#'
#'
#' The traditionaly, the so-called FPF;\emph{False Positive Fraction} and TPT:\emph{True Positive Fraction} are used.
#' Recall that our data format:
#'
#'
#'
#'
#' \emph{ A single reader and a single modality case }
#'
#'------------------------------------------------------------------------------------------------------
#' \tabular{llll}{
#' \code{NI, NL } \tab \strong{ confidence level } \tab \strong{ No. of false alarms} \tab \strong{No. of hits} \cr
#' \tab \tab (FP:False Positive) \tab (TP:True Positive) \cr
#' -----------------------\tab ----------------------- \tab ----------------------------- \tab ------------- \cr
#' \emph{definitely} present \tab 5 \tab \eqn{F_5} \tab \eqn{H_5} \cr
#' \emph{probably} present \tab 4 \tab \eqn{F_4} \tab \eqn{H_4} \cr
#' equivocal \tab 3 \tab \eqn{F_3} \tab \eqn{H_3} \cr
#' subtle \tab 2 \tab \eqn{F_2} \tab \eqn{H_2} \cr
#' \emph{very} subtle \tab 1 \tab \eqn{F_1} \tab \eqn{H_1} \cr
#' }
#'
#'---------------------------------------------------------------------------------------------------
#'
#' FPF is defined as follows;
#'
#'
#' \deqn{FPF(5):= \frac{F_5}{NI},}
#' \deqn{FPF(4):= \frac{F_4+F_5}{NI},}
#' \deqn{FPF(3):= \frac{F_3+F_4+F_5}{NI},}
#' \deqn{FPF(2):= \frac{F_2+F_3+F_4+F_5}{NI},}
#' \deqn{FPF(1):= \frac{F_1+F_2+F_3+F_4+F_5}{NI}.}
#'
#'
#' TPF is defined as follows;
#'
#'
#' \deqn{TPF(5):= \frac{H_5}{NL},}
#' \deqn{TPF(4):= \frac{H_4+H_5}{NL},}
#' \deqn{TPF(3):= \frac{H_3+H_4+H_5}{NL},}
#' \deqn{TPF(2):= \frac{H_2+H_3+H_4+H_5}{NL},}
#' \deqn{TPF(1):= \frac{H_1+H_2+H_3+H_4+H_5}{NL}.}
#'
#'
#'
#'
#'
#'
#'
#'
#'@param dataList A list, consisting of the following \R objects:\code{m,q,c,h,f,NL,C,M,Q} each of which means from the right
#'
#'\code{caseID } : A vector, indicating the case ID (image, radiograph, patient ... etc) = 1,2,... which does not include zero.
#'
#'\code{m } : A vector, indicating the modality ID = 1,2,... which does not include zero.
#'
#'\code{q } : A vector, indicating the reader ID = 1,2,... which does not include zero.
#'
#'\code{c } : A vector, indicating the confidence = 1,2,... which does not include zero.
#'
#'\code{h } : A vector, indicating the number of hits
#'
#'\code{f } : A vector, indicating the number of false alarm
#'
#'\code{NL } : A positive integer, indicating the number of lesions for all images
#'
#'\code{C } : A positive integer, indicating the highest number of confidence level
#'
#'\code{M } : A positive integer, indicating the number of modalities
#'
#'\code{Q } : A positive integer, indicating the number of readers.
#'
#'The detail of these dataset, please see the example datasets, e.g. \code{\link{dd}}.
# @return-----
#'@return A list, which includes arrays and vectors.
#'A metadata such as number of cumulative false alarms
#' and hits to create and draw the curve.
#'
#' The
#' \emph{False Positive Fraction (FPF)} and
#' \emph{True Positive Fraction (TPF)} are also calculated.
#'
#'\strong{ The components of list} \emph{ I rediscover it at 2019 Jun 18, I am not sure it is useful? 2019 Dec 8}
#' \describe{
#'\item{ \code{ harray} }{An array of hit, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ farray} }{An array of false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharray} }{An array of \strong{cumulative} hits, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarray} }{An array of \strong{cumulative} false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharrayN} }{An array of TPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarrayN} }{An array of FPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ h} }{An vector of hit, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ f } }{An vector of false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hh } }{An vector of \strong{cumulative} hits, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ff } }{An vector of \strong{cumulative} false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hhN} }{An vector of TPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffN} }{An vector of FPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'}
#'
#'
#' Revised Nov. 21
#'
#'
#'
#'@examples
#'#========================================================================================
#'# First, we prepare the data endowed with this package.
#'#========================================================================================
#'
#'
#'
#' dat <- get(data("dataList.Chakra.Web"))
#'
#'
#'#========================================================================================
#'# #Calculate FPFs and TPFs and etc.
#'#========================================================================================
#'
#'
#'
#' a <- metadata_to_fit_MRMC(dat)
#'
#'
#' #Now, we get a meta-data object named "a".
#'
#'
#'#========================================================================================
#'# Check of Definiion
#'#========================================================================================
#'
#'
#' a$hh/dat$NL == a$hhN
#'
#'# Since all of aboves are TRUE, the hhN is a TPF per NL.
#'
#'
#'
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs
#'#========================================================================================
#'#'
#'
#'
#'
#' FPF = a$ffN
#' TPF = a$hhN
#'
#' dark_theme()
#' plot(FPF,TPF)
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs via ggplot
#'#========================================================================================
#'
#' length(dat$f)==length(FPF)
#'
#' q <- dat$q
#' m <- dat$m
#' df <- data.frame(FPF,
#' TPF,
#' m,
#' q
#' )
#'
#'# ggplot2::ggplot(df, aes(x =FPF, y = TPF, colour = q, group = m)) + ggplot2::geom_point()
#'
#'# Revised 2019 Jun 18, Revised 2019 Sept 9
#'
#'
#'
#'
#'
# devtools::document();help("metadata_MRMC") # Confirm reflection
# devtools::use_data(dataList.high.ability)
#' @export metadata_to_fit_MRMC
#'@inheritParams fit_Bayesian_FROC
metadata_to_fit_MRMC_casewise<- function(dataList,ModifiedPoisson=FALSE)
{
# message("\n")
# message("* Now, we calculated the metadata, e.g., cumulative hits and false alarms etc... \n")
NL_casewise <- dataList$NL_casewise
caseIDs_deseased <- dataList$caseIDs_deseased
NI_deseased <- length( caseIDs_deseased)
caseID <- dataList$caseID
m <-dataList$m
q<-dataList$q
c<-dataList$c
h<-dataList$h
f<-dataList$f
NI<-dataList$NI
NL<-dataList$NL
C<-dataList$C
M<-dataList$M
Q<-dataList$Q
modalityID_dummy <- fastDummies::dummy_cols( data.frame( modalityID_dummy=m) )
modalityID_dummy <- modalityID_dummy[, colnames(modalityID_dummy) != "modalityID_dummy"]
modalityID_dummy <- as.matrix(modalityID_dummy)
readerID_dummy <- fastDummies::dummy_cols( data.frame( readerID_dummy=q) )
readerID_dummy <- readerID_dummy[, colnames(readerID_dummy) != "readerID_dummy"]
readerID_dummy <- as.matrix(readerID_dummy)
caseID_dummy <- fastDummies::dummy_cols( data.frame( caseID_dummy=caseID) )
caseID_dummy <- caseID_dummy[, colnames(caseID_dummy) != "caseID_dummy"]
caseID_dummy <- as.matrix(caseID_dummy)
if(ModifiedPoisson==FALSE) NX = NI;
if(ModifiedPoisson==TRUE) NX =NL;
N <- NI*M*Q*C
hh <- numeric(N) #Initialization of Cumulative Hits
ff <- numeric(N) #Initialization of Cumulative False alarm
hharray<-array(0,dim=c(NI,C,M,Q));#Cumulative sum
ffarray<-array(0,dim=c(NI,C,M,Q));#Cumulative sum
harray<-array(0,dim=c(NI,C,M,Q));#Non-Cumulative
farray<-array(0,dim=c(NI,C,M,Q)); #Non-Cumulative
# hCQ<-array(0,dim=c(C,Q));#Non-Cumulative, that is, merely format is only changed from vector to array.
# fCQ<-array(0,dim=c(C,Q)); #Non-Cumulative, that is, merely format is only changed from vector to array.
# hC<-array(0,dim=c(C));#Non-Cumulative, that is, merely format is only changed from vector to array.
# fC<-array(0,dim=c(C)); #Non-Cumulative, that is, merely format is only changed from vector to array.
for(case in 1:NI){
for(mmm in 1:M) {
for(ccc in 1:C) {
for(qqq in 1 : Q){
for(n in 1:ccc){
ff[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]<-ff[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]+f[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-n)]
hh[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]<-hh[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]+h[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-n)]
}
ffarray[case,ccc,mmm,qqq] <-ff[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
hharray[case,ccc,mmm,qqq] <-hh[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
farray[case,ccc,mmm,qqq] <- f[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
harray[case,ccc,mmm,qqq] <- h[(case -1 )*M*Q*C + (mmm-1)*Q*C+(qqq-1)*C+(C+1-ccc)]
}}}}
# for(cd in 1:C){
# for(qd in 1:Q){
# for(md in 1:M){
# # hCQ[cd, qd]<-hCQ[cd, qd]+harray[cd,md,qd]
# # fCQ[cd, qd]<-fCQ[cd, qd]+farray[cd,md,qd]
# }}}
# for(cd in 1:C){
# for(qd in 1:Q){
# for(md in 1:M){
# # hC[cd]<-hC[cd]+harray[cd,md,qd]
# # fC[cd]<-fC[cd]+farray[cd,md,qd]
# }}}
# S <-array(0,dim=c(M,Q));
# for(md in 1:M){
# for(qd in 1 : Q){
# S[md,qd] <- sum(harray[,md,qd])
# }}
hharrayN<-hharray/NL
ffarrayN<-ffarray/NX
hhN<-hh/NL
ffN<-ff/NX
data <- list(N=N,Q=Q, M=M,m=m ,C=C ,caseID=caseID,NI_deseased=NI_deseased,
# S=S,
NL=NL,NI=NI, ccc = C:1
,c=c,q=q,
h=h, f=f,
hh=hh, hhN=hhN,
ff=ff,ffN=ffN,
harray=harray, farray=farray,
hharray=hharray, ffarray=ffarray,
hharrayN=hharrayN, ffarrayN=ffarrayN,
NL_casewise=NL_casewise,
caseIDs_deseased=caseIDs_deseased,
# Return ----
modalityID_dummy = modalityID_dummy,
readerID_dummy = readerID_dummy,
caseID_dummy = caseID_dummy
)
return(data)
}
#' @title Create metadata for MRMC data.
#'
#'@description
#' The so-called \emph{false positive fraction (FPF)}
#' and the \emph{true positive fraction (TPF)}
#' are calculated from the number of hits (True Positives: TPs)
#' and the number of false alarms (False Positives: FPs)
#'
#'
#'
#'
#'
#'
#'
#'@details
#' From data of number of hits (True Positive: TP)
#' and false alarms (False Positive: FP), we calculate the number
#' of cumulative false positives (FPF) and cumulative hits (TPF).
#'
#' Because there are three subscripts,
#' reader, modality, and image,
#' we create array format and vector format etc...
#'
#'@param dataList A list, should include
#' \code{m,q,c,h,f,NL,C,M,Q} which means
#'
#'\code{c} should be
#'created by \code{ c <-c(rep(C:1))},
#' where \code{C} is the number of confidence levels.
#'So, you should write down your hits
#'and false alarms vector so that it
#'is compatible with this
#' automatically created \code{c} vector.
#'
#'\code{h} means the number of hits
#'
#'\code{f} means the number of false alarm
#'
#'\code{NL} means the Total number of lesions for all images
#'
#'\code{C} means the highest number of confidence level
#'
#'@return A metadata such as number
#' of cumulative false alarms and
#' hits to create and draw the curve.
#'
#'
#'
#'
#'
#'
#'
#'@examples
#' \dontrun{
#'#========================================================================================
#'# TP and FP
#'#========================================================================================
#'
#'
#' dat <- BayesianFROC::dataList.Chakra.Web
#'
#'
#'
#'#========================================================================================
#'# Calculates TPF and FPF from TP and FP
#'#========================================================================================
#'
#'
#' metadata_srsc_per_image(dat)
#'
#'
#'
#'
#'
#'
#' # Revised 2019 Nov.
#'
#'
#'}# dottest
#' @export metadata_srsc_per_image
#'@inheritParams fit_Bayesian_FROC
metadata_srsc_per_image<- function(dataList,ModifiedPoisson)
{
if ( length(dataList)>=7)return(message("srsc Only. Your data is not srsc."))# This cause error in ReadMe, so It is better to omit this line. 2019 Sept
h <- dataList$h
f <- dataList$f
NI <- dataList$NI
NL <- dataList$NL
C <- as.integer(dataList$C)
#
# if(ModifiedPoisson)NX <- NL
# if(!ModifiedPoisson)NX <- NI
#
NX <- if(ModifiedPoisson) NL else if(!ModifiedPoisson) NI
# browser()
# C <-length(h)
c <-c(rep(C:1))
if ( (length(h) >length(f))
|| (length(h) <length(f))
) { return(message("Format error:\nIn your data, true positive and false positives are not same length.\n"))
} else
if ( (sum(h, na.rm = TRUE) >NL) )
{return(message("Format error:\nIn your data, number of true positives are greater than that of lesions.\n"))
} else
N <- length(f)
# hh <- numeric(N) #cumulative fraction
# ff <- numeric(N) #cumulative fraction
# for(cd in 1:C) {
# for(n in 1:cd) {
# hh[cd]<-hh[cd]+h[n]/NL #cumulative fraction to examine the fitness of FROC
# ff[cd]<-ff[cd]+f[n]/NI #cumulative fraction to examine the fitness of FROC
# }}
# fff <- numeric(N)
# for(cd in 1:C) {
# for(n in 1:cd) {
# fff[cd]<-fff[cd]+f[n] #cumulative fraction to examine the fitness of FROC
# }}
fff <- cumsum(f)
hh <- cumsum( h)/NL
ff <- fff/NX
hitExtended <- vector( length =(C+1),mode = "integer")
falsealarmExtended <- vector( length =(C+1),mode = "integer")
for (cd in 1:C) hitExtended[cd] = h[cd];
hitExtended[C+1]=NL-sum(h, na.rm = TRUE);
for (cd in 1:C) falsealarmExtended[cd] = f[cd];#2020 Dec 7
falsealarmExtended[C+1]=NI-sum(f);#2020 Dec 7
data <- list( N=N,NL=NL,NI=NI,C=as.integer(C),c=c,
h=h,f=f,
hh=hh,ff=ff,
fff=fff,
hitExtended = hitExtended,
falsealarmExtended=falsealarmExtended
)
invisible(data)
}
# title -----
#' @title Create metadata for MRMC data
#'
#'
#'@description
#' The so-called \emph{false positive fraction (FPF)}
#' and the \emph{true positive fraction (TPF)}
#' are calculated from the number of hits (True Positives: TPs)
#' and the number of false alarms (False Positives: FPs)
#'
#'
#'
#'
#'
#'
#'
#'
#'
#'@details To fit a model to data,
#' we need a hit data and false data
#' formulated by both an array and a vector.
#'
#' It also calculates the so-called
#' False Positive Fractions (FPF)
#' (resp. True Positive Fractions (TPF) )
#' which are cumulative sums of false alarms (resp.
#' hits) over number of lesions or images.
#'
#' From data of number of hits
#' and false alarms,
#' we calculate the number
#' of cumulative false positives
#' and hits per image or lesion, in other words,
#' \emph{False Positive Fraction (FPF)}
#' and \emph{True Positive Fraction (TPF)}.
#' Since there are three subscripts,
#' \emph{reader}, \emph{modality},
#' and \emph{image}, we can create array format
#' or vector format etc...
#'
#' \strong{Abbreviations}
#'
#' \emph{FPF: false positive fraction }
#'
#' \emph{TPF: true positive fraction }
#'
#' \emph{hit : True Positive = TP}
#'
#' \emph{false alarms: False Positive = FP}
#'
#'
#'
#'
#'
#'
#' The traditionaly, the so-called FPF;\emph{False Positive Fraction} and TPT:\emph{True Positive Fraction} are used.
#' Recall that our data format:
#'
#'
#'
#'
#' \emph{ A single reader and a single modality case }
#'
#'------------------------------------------------------------------------------------------------------
#' \tabular{llll}{
#' \code{NI, NL } \tab \strong{ confidence level } \tab \strong{ No. of false alarms} \tab \strong{No. of hits} \cr
#' \tab \tab (FP:False Positive) \tab (TP:True Positive) \cr
#' -----------------------\tab ----------------------- \tab ----------------------------- \tab ------------- \cr
#' \emph{definitely} present \tab 5 \tab \eqn{F_5} \tab \eqn{H_5} \cr
#' \emph{probably} present \tab 4 \tab \eqn{F_4} \tab \eqn{H_4} \cr
#' equivocal \tab 3 \tab \eqn{F_3} \tab \eqn{H_3} \cr
#' subtle \tab 2 \tab \eqn{F_2} \tab \eqn{H_2} \cr
#' \emph{very} subtle \tab 1 \tab \eqn{F_1} \tab \eqn{H_1} \cr
#' }
#'
#'---------------------------------------------------------------------------------------------------
#'
#' FPF is defined as follows;
#'
#'
#' \deqn{FPF(5):= \frac{F_5}{NI},}
#' \deqn{FPF(4):= \frac{F_4+F_5}{NI},}
#' \deqn{FPF(3):= \frac{F_3+F_4+F_5}{NI},}
#' \deqn{FPF(2):= \frac{F_2+F_3+F_4+F_5}{NI},}
#' \deqn{FPF(1):= \frac{F_1+F_2+F_3+F_4+F_5}{NI}.}
#'
#'
#' TPF is defined as follows;
#'
#'
#' \deqn{TPF(5):= \frac{H_5}{NL},}
#' \deqn{TPF(4):= \frac{H_4+H_5}{NL},}
#' \deqn{TPF(3):= \frac{H_3+H_4+H_5}{NL},}
#' \deqn{TPF(2):= \frac{H_2+H_3+H_4+H_5}{NL},}
#' \deqn{TPF(1):= \frac{H_1+H_2+H_3+H_4+H_5}{NL}.}
#'
#'
#'
#'
#'
#'
#'
#'
#'@param dataList A list, consisting of the following \R objects:\code{m,q,c,h,f,NL,C,M,Q} each of which means from the right
#'
#'\code{m } : A vector, indicating the modality ID = 1,2,... which does not include zero.
#'
#'\code{q } : A vector, indicating the reader ID = 1,2,... which does not include zero.
#'
#'\code{c } : A vector, indicating the confidence = 1,2,... which does not include zero.
#'
#'\code{h } : A vector, indicating the number of hits
#'
#'\code{f } : A vector, indicating the number of false alarm
#'
#'\code{NL } : A positive integer, indicating the number of lesions for all images
#'
#'\code{C } : A positive integer, indicating the highest number of confidence level
#'
#'\code{M } : A positive integer, indicating the number of modalities
#'
#'\code{Q } : A positive integer, indicating the number of readers.
#'
#'The detail of these dataset, please see the example datasets, e.g. \code{\link{dd}}.
# @return-----
#'@return A list, which includes arrays and vectors.
#'A metadata such as number of cumulative false alarms
#' and hits to create and draw the curve.
#'
#' The
#' \emph{False Positive Fraction (FPF)} and
#' \emph{True Positive Fraction (TPF)} are also calculated.
#'
#'\strong{ The components of list} \emph{ I rediscover it at 2019 Jun 18, I am not sure it is useful? 2019 Dec 8}
#' \describe{
#'\item{ \code{ harray} }{An array of hit, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ farray} }{An array of false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharray} }{An array of \strong{cumulative} hits, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarray} }{An array of \strong{cumulative} false alarms, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hharrayN} }{An array of TPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffarrayN} }{An array of FPF, dimension \code{ [C,M,Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ h} }{An vector of hit, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ f } }{An vector of false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hh } }{An vector of \strong{cumulative} hits, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ff } }{An vector of \strong{cumulative} false alarms, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ hhN} }{An vector of TPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'\item{ \code{ ffN} }{An vector of FPF, dimension \code{ [C*M*Q]}, where \code{C,M,Q} are a number of confidence level, modalities, readers, respectively. }
#'}
#'
#'
#' Revised Nov. 21
#'
#'
#'
#'@examples
#'#========================================================================================
#'# First, we prepare the data endowed with this package.
#'#========================================================================================
#'
#'
#'
#' dat <- get(data("dataList.Chakra.Web"))
#'
#'
#'#========================================================================================
#'# #Calculate FPFs and TPFs and etc.
#'#========================================================================================
#'
#'
#'
#' a <- metadata_to_fit_MRMC(dat)
#'
#'
#' #Now, we get a meta-data object named "a".
#'
#'
#'#========================================================================================
#'# Check of Definiion
#'#========================================================================================
#'
#'
#' a$hh/dat$NL == a$hhN
#'
#'# Since all of aboves are TRUE, the hhN is a TPF per NL.
#'
#'
#'
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs
#'#========================================================================================
#'#'
#'
#'
#'
#' FPF = a$ffN
#' TPF = a$hhN
#'
#' dark_theme()
#' plot(FPF,TPF)
#'
#'#========================================================================================
#'# Plot a FPFs and TPFs via ggplot
#'#========================================================================================
#'
#' length(dat$f)==length(FPF)
#'
#' q <- dat$q
#' m <- dat$m
#' df <- data.frame(FPF,
#' TPF,
#' m,
#' q
#' )
#'
#'# ggplot2::ggplot(df, aes(x =FPF, y = TPF, colour = q, group = m)) + ggplot2::geom_point()
#'
#'# Revised 2019 Jun 18, Revised 2019 Sept 9
#'
#'
#'
#'
#'
# devtools::document();help("metadata_MRMC") # Confirm reflection
# devtools::use_data(dataList.high.ability)
#' @export metadata_to_fit_MRMC
#'@inheritParams fit_Bayesian_FROC
metadata_to_fit_MRMC<- function(dataList,ModifiedPoisson=FALSE)
{
# message("\n")
# message("* Now, we calculated the metadata, e.g., cumulative hits and false alarms etc... \n")
m <-dataList$m
q<-dataList$q
c<-dataList$c
h<-dataList$h
f<-dataList$f
NI<-dataList$NI
NL<-dataList$NL
C<-dataList$C
M<-dataList$M
Q<-dataList$Q
if(ModifiedPoisson==FALSE) NX = NI;
if(ModifiedPoisson==TRUE) NX =NL;
N <-C*M*Q
#For Draw the Sample points on FROC curve. Assessment of Fit for FROC.
hh <- numeric(N) #Initialization of Cumulative Hits
ff <- numeric(N) #Initialization of Cumulative False alarm
hharray<-array(0,dim=c(C,M,Q));#Cumulative
ffarray<-array(0,dim=c(C,M,Q));#Cumulative
harray<-array(0,dim=c(C,M,Q));#Non-Cumulative
farray<-array(0,dim=c(C,M,Q)); #Non-Cumulative
hCQ<-array(0,dim=c(C,Q));#Non-Cumulative, that is, merely format is only changed from vector to array.
fCQ<-array(0,dim=c(C,Q)); #Non-Cumulative, that is, merely format is only changed from vector to array.
hC<-array(0,dim=c(C));#Non-Cumulative, that is, merely format is only changed from vector to array.
fC<-array(0,dim=c(C)); #Non-Cumulative, that is, merely format is only changed from vector to array.
for(md in 1:M) {
for(cd in 1:C) {
for(qd in 1 : Q){
for(n in 1:cd){
ff[cd+(md-1)*C*Q+(qd-1)*C]<-ff[cd+(md-1)*C*Q+(qd-1)*C]+f[n+(md-1)*C*Q+(qd-1)*C]
hh[cd+(md-1)*C*Q+(qd-1)*C]<-hh[cd+(md-1)*C*Q+(qd-1)*C]+h[n+(md-1)*C*Q+(qd-1)*C]
}
ffarray[cd,md,qd] <-ff[cd+(md-1)*C*Q+(qd-1)*C]
hharray[cd,md,qd] <-hh[cd+(md-1)*C*Q+(qd-1)*C]
farray[cd,md,qd] <- f[cd+(md-1)*C*Q+(qd-1)*C]
harray[cd,md,qd] <- h[cd+(md-1)*C*Q+(qd-1)*C]
}}}
for(cd in 1:C){
for(qd in 1:Q){
for(md in 1:M){
hCQ[cd, qd]<-hCQ[cd, qd]+harray[cd,md,qd]
fCQ[cd, qd]<-fCQ[cd, qd]+farray[cd,md,qd]
}}}
for(cd in 1:C){
for(qd in 1:Q){
for(md in 1:M){
hC[cd]<-hC[cd]+harray[cd,md,qd]
fC[cd]<-fC[cd]+farray[cd,md,qd]
}}}
S <-array(0,dim=c(M,Q));
for(md in 1:M){
for(qd in 1 : Q){
S[md,qd] <- sum(harray[,md,qd])
}}
hharrayN<-hharray/NL
ffarrayN<-ffarray/NX
hhN<-hh/NL
ffN<-ff/NX
data <- list(N=N,Q=Q, M=M,m=m ,C=C ,S=S, NL=NL,NI=NI, ccc = C:1
,c=c,q=q,
h=h, f=f,
hh=hh, hhN=hhN,
ff=ff,ffN=ffN,
harray=harray, farray=farray,
hharray=hharray, ffarray=ffarray,
hharrayN=hharrayN, ffarrayN=ffarrayN)
return(data)
}
#' @title Create metadata for MRMC data
#'@description From data of number of hits and false alarms, we calculate the number of cumulative false positives and hits.
#' Since there are three subscripts, reader, modality, and image, we create array format and vector format etc...
#'@inheritParams fit_Bayesian_FROC
#'@inheritParams DrawCurves
#'@return A metadata such as number of cumulative false alarms and hits to create and draw the curve.
#' @export metadata_to_DrawCurve_MRMC
metadata_to_DrawCurve_MRMC<- function(
StanS4class,
mesh.for.drawing.curve=5000)
{
fit <-StanS4class
data <-fit@metadata
m<-data$m ;S<-data$S; NL<-data$NL;c<-data$c;q<-data$q;
h<-data$h; f<-data$f;
hh<-data$hh; hhN<-data$hhN;
ff<-data$ff;ffN<-data$ffN;
harray<-data$harray; farray<-data$farray;
hharray<-data$hharray; ffarray<-data$ffarray;
hharrayN<-data$hharrayN; ffarrayN<-data$ffarrayN;
C<-as.integer(data$C)
M<-as.integer(data$M)
N<-as.integer(data$N)
Q<-as.integer(data$Q)
# message(crayon::silver("\n* Metadata to draw the curves are callculating ... \n"))
fit <- methods::as(StanS4class, "stanfit")
war <- fit@sim$warmup
cha <- fit@sim$chains
ite <- fit@sim$iter
#Draw the AFROC curve~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
MCMC=(ite-war)*cha #Total Samples By MCMC
ll<- stats::rchisq(mesh.for.drawing.curve, 1)
lll<- 0.99+stats::rchisq(mesh.for.drawing.curve, 1)
l<-append(ll,lll) #This name l is duplicated!! CAUTION!!
#sort ----
l<-sort(l, method = "shell", index.return = FALSE)
x<- 1-exp(-l) #AFROC-Curve
y <- array(0, dim=c(length(x), M,Q)) #
# a<-rstan::extract(fit)$a #samples of a by MCMC
# b<-rstan::extract(fit)$b #samples of b by MCMC
# EAP_a <- array(0, dim=c( M,Q)) #
# EAP_b <- array(0, dim=c( M,Q)) #
#
# for(md in 1:M){
# for(qd in 1:Q){
#
# EAP_a[md,qd] <- 0
# EAP_b[md,qd] <- 0
# s<-0
# t<-0
# for(mc in 1:MCMC){ #EAP
# s<- EAP_a[md,qd]
# EAP_a[md,qd] <- s+ a[mc,md,qd]
# t<- EAP_b[md,qd]
# EAP_b[md,qd] <- t+ b[mc,md,qd]
# }
# EAP_a[md,qd] <-EAP_a[md,qd] /MCMC #EAP of a
# EAP_b[md,qd] <-EAP_b[md,qd] /MCMC #EAP of b
# }}
EAP_a <- extract_EAP_by_array(fit,"aa")
EAP_b <- extract_EAP_by_array(fit,"bb")
EAP_AA <- extract_EAP_by_array(fit,"AA")
# message("* Process for calculating y coordinates of curves \n")#Processsssssss
plotFROCdata <- list()
plotAFROCdata <- list()
nameFROC <- vector()
nameAFROC <- vector()
#Vectorization #Vectorization #Vectorization #Vectorization
for(i in 1:length(x)){ #y-coord.of FROC and AFROC are same
y[i,,]<-1-stats::pnorm(EAP_b *stats::qnorm(exp(-l[i]))-EAP_a )#20210120
}
#Vectorization #Vectorization #Vectorization #Vectorization
s <- 0
for(qd in 1:Q){
for(md in 1:M){
s<-s+1
# message("|")#Processsssssss
plotFROCdata[[s]] <- data.frame(x.coord = l, y.coord =y[,md,qd] )
plotAFROCdata[[s]] <- data.frame(x.coord = x, y.coord =y[,md,qd] )
names( plotFROCdata[[s]] ) <- c(paste("X.coord.for.FROC.curve.for.modality.",md,".and.reader.",qd,sep = ""),paste("Y.coord.for.FROC.curve.for.modality.",md,".and.reader.",qd,sep = ""))
names( plotAFROCdata[[s]] ) <- c(paste("X.coord.for.AFROC.curve.for.modality.",md,".and.reader.",qd,sep = ""),paste("Y.coord.for.AFROC.curve.for.modality.",md,".and.reader.",qd,sep = ""))
nameFROC[s] <- paste("FROC.data.for.modality.",md,".and.reader.",qd,sep = "")
nameAFROC[s] <- paste("AFROC.data.for.modality.",md,".and.reader.",qd,sep = "")
# names(list$plotAFROCdata)
}
# message(paste("", ceiling(round(qd/Q,2)*100/2),"% \n"))#Processsssssss
}
names(plotFROCdata) <-nameFROC
names(plotAFROCdata) <-nameAFROC
#aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
#-- AUC for each modarlity and each reader
# AA<-rstan::extract(fit)$AA #samples of a by MCMC
# EAP_AA <- array(0, dim=c( M,Q)) #
#
# for(qd in 1:Q){
# for(md in 1:M){
# EAP_AA[md,qd] <- mean(AA[,md,qd])
# # EAP_AA[md,qd] <- 0
# # s<-0
# # message("|")#Processsssssss
# # for(mc in 1:MCMC){ #EAP
# # s<- EAP_AA[md,qd]
# # EAP_AA[md,qd] <- s+ AA[mc,md,qd]
# # }
# # EAP_AA[md,qd] <-EAP_AA[md,qd] /MCMC #EAP of a
# }#for md
# # message(paste("", ceiling(round(qd/Q,2)*100/2 +50),"% \n"))#Processsssssss
# }# for qd
# aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
# #--------- chi ^2 -----------Start
# p<-rstan::extract(fit)$p
# lambda<-rstan::extract(fit)$l
# EAP_p <- array(0, dim=c( C,M,Q))
# EAP_l <- array(0, dim=c( C+1))
#
# s <- array(0, dim=c( C,M,Q))
# t <- array(0, dim=c( C+1))
# for(mc in 1:MCMC){
# for(cd in 1:C){
# for(md in 1:M){
# for(qd in 1:Q){
# s[ cd,md,qd]<- EAP_p[ cd,md,qd]
#
# EAP_p[ cd,md,qd] <- s[ cd,md,qd]+ p[mc,cd,md,qd]
# }
# for(cd in 0:C){
# t[ cd]<- EAP_l[ cd]
# EAP_l[ cd] <- t[ cd]+ lambda[mc,cd]
# }
# }}}
# EAP_p<-EAP_p/MCMC
# EAP_l<-EAP_l/MCMC
#
# ss<-array(0, dim=c( M,Q))
# tt<-array(0, dim=c( M,Q))
# for(cd in 1:C){
# for(md in 1:M){
# for(qd in 1:Q){
# ss[md,qd]<-(hharray[C+1-cd,md,qd]-NL*EAP_p[cd,md,qd])^2/(NL*EAP_p[cd,md,qd])
# ## tt<-(ffarray[C+1-cd,md,qd]-NI*(EAP_l[cd ]-EAP_l[cd+1 ]))^2/(NI*(EAP_l[cd ]-EAP_l[cd+1 ]))
# tt[md,qd]<-(ffarray[C+1-cd,md,qd]-NL*(EAP_l[cd ]-EAP_l[cd+1 ]))^2/(NL*(EAP_l[cd ]-EAP_l[cd+1 ]))
#
# }}}
# chisquare <- ss+tt
# chisquare
#
invisible(list(x=x,y=y,l=l,EAP_AA=EAP_AA ,plotFROCdata=plotFROCdata,plotAFROCdata=plotAFROCdata))
}
Try the BayesianFROC package in your browser
Any scripts or data that you put into this service are public.
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.