Nothing
R/BOP2_functions_EffTox.R
In GBOP2: Generalized Bayesian Optimal Phase II Design (G-BOP2)
Defines functions
DoStoppingBoundaries_minN
DoStoppingBoundaries
ncore <- 1
# ------------------------
DoStoppingBoundaries = function( input, globalPara, Calc, nobs.seq, nobs.seqTX, isSBcalculated, session )
{
#browser();
if(input$e1n>0.0001 & input$e1a>0.0001 & input$e1n<1 & input$e1a<1 &
input$e2n>0.0001 & input$e2a>0.0001 & input$e2n<1 & input$e2a<1 &
input$e3n>0.0001 & input$e3a>0.0001 & input$e3n<1 & input$e3a<1 &
!is.na(input$e1n) & !is.na(input$e1a) & !is.na(input$e2n) & !is.na(input$e2a) & !is.na(input$e3n) & !is.na(input$e3a)){
if(input$e1n<=input$e1a & input$e2n>=input$e2a) {
if(input$e1n+input$e2n-input$e3n<1 & input$e1a+input$e2a-input$e3a<1 &
input$e3n<min(input$e1n,input$e2n) & input$e3a<min(input$e1a,input$e2a)){
if((input$ETprior1>=0 & input$ETprior1<=1 & input$ETprior2>=0 & input$ETprior2<=1 & input$ETprior3>=0 & input$ETprior3<=1 &
!is.na(input$ETprior1) & !is.na(input$ETprior2) & !is.na(input$ETprior3) & !is.na(input$ETpriorSS) & input$ETpriorSS>0)| input$priorspec){
if(input$ETprior1+input$ETprior2-input$ETprior3 <= 1 & input$ETprior3 <= min(input$ETprior1,input$ETprior2)){
# cut.seq = sort(c(seq(0.5,0.95,by=0.025),seq(0.96,0.99,0.01),seq(0.9-input$err1,1-input$err1,0.005)))
cut.seq = rev(c(seq(0.5,0.8,0.025),seq(0.81,0.96,0.01)))
power.seq = log(seq(1,0.5,by=-0.025))/log(0.5)
p.n = c(input$e3n,input$e1n-input$e3n,input$e2n-input$e3n,1-input$e1n-input$e2n+input$e3n)
##input$e3n: Pr(Eff & Tox), input$e1n: Pr(Eff), input$e2n: Pr(Tox)
##input$e1n-input$e3n: Pr(Eff & no Tox)
##input$e2n-input$e3n: Pr(no Eff & Tox)
##1-input$e1n-input$e2n+input$e3n: Pr(no Eff & no Tox)
p.a = c(input$e3a,input$e1a-input$e3a,input$e2a-input$e3a,1-input$e1a-input$e2a+input$e3a)
# if(input$priorspec){
# prior = p.n
# }else{
# prior = c(input$ETprior3*input$ETpriorSS,(input$ETprior1-input$ETprior3)*input$ETpriorSS,
# (input$ETprior2-input$ETprior3)*input$ETpriorSS,(1-input$ETprior1-input$ETprior2+input$ETprior3)*input$ETpriorSS)
# }
oc.mat.efftox = c()
cut.start = 1
# oc.mat.efftox = Calc$GridSearchToxEffRcpp(seed=input$seed, contrast = matrix(c(1,1,0,0,0,1,0,1),nrow=2,byrow=TRUE),
# nobs = nobs.seq, nobsTX = nobs.seqTX, dprior = p.n,
# b = cut.seq, pow = power.seq, pn=p.n, pa = p.a,
# phi=c(input$e1n,1-input$e2n), cutstart=cut.start,
# nsim = input$numOfSimForTiralSetting,
# err1 = input$err1, maxresp,resp_list_r)
########################################################
r0 = input$e1n
t0 = input$e2n
r1 = input$e1a
t1 = input$e2a
interm.eff = nobs.seq #
interm.tox = nobs.seqTX #
message(unique(sort(c(interm.eff, interm.tox)))) #
scen = hypotheses_ind(r0, r1, t0, t1) ##independent
##########################################################
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
oc.mat.efftox.temp= grid_search_cpp(interm.eff=interm.eff,
interm.tox=interm.tox,
scenario=scen,
r0 = r0, r1= r1,
t0=t0, t1= t1,
lambda_r=cut.seq,
lambda_t= cut.seq,
gamma_space=power.seq,
typeI01=input$err_eff, typeI10=input$err_tox,
typeI00=input$err_all,
true_opt = input$truet1e,
prior = scen[1,],
cppfunction=GridSearchBiRcpp) # return a vector
if(input$e3n == (input$e1n*input$e2n)& input$e3a == (input$e1a*input$e2a) ){
oc.mat.efftox =oc.mat.efftox.temp
message("Independent")
}else{
message("Not independent")
optim.temp=oc.mat.efftox.temp[1:3]
lambda_E_space = seq(optim.temp[1]-0.2, min(1,optim.temp[1]+0.05 ),by=0.01)
lambda_T_space = seq(optim.temp[2]-0.2, min(1,optim.temp[2]+0.05 ),by=0.01)
gamma_space = seq(optim.temp[3]-0.2, min(1,optim.temp[3]+0.1 ),by=0.02)
scen = hypotheses_corr(r0,r1,t0,t1,PA_ET=input$e3a, PN_ET = input$e3n)
message(scen)
# joint probabilities of four possible outcomes
# (no efficacy and no toxicity, no efficacy and toxicity, efficacy and no toxicity, both efficacy and toxicity)
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
search_space = bond.unique(interm.eff=interm.eff, interm.tox=interm.tox, scenario=scen,
lambda_r = lambda_E_space,
lambda_t = lambda_T_space,
gamma_space= gamma_space,
r0=r0,t0=t0,
prior = scen[1,])
oc.mat.efftox=optimization_corr(interm.eff=interm.eff,
interm.tox=interm.tox,
search.temp = search_space,
scenario=scen,
r0=r0, t0=t0,
typeI01=input$err_eff,
typeI10=input$err_tox,
typeI00=input$err_all,
lpow=0.3,
ncore=1)
}
if(is.null(oc.mat.efftox) | length(oc.mat.efftox)==0) {
globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$boundary=c()
globalPara$prior=c()
} else {
optim=oc.mat.efftox
#message(oc.mat.efftox[1:10,])
# oc.mat.efftoxS = oc.mat.efftox[oc.mat.efftox[,3]<=input$err1,]
# oc.mat.efftoxL = oc.mat.efftox[oc.mat.efftox[,3]>input$err1,]
#
# #message(oc.mat.efftoxS)
# #message(oc.mat.efftoxL)
#
# if(is.null(oc.mat.efftoxS) | length(oc.mat.efftoxS)==0) {
# globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
# globalPara$calculated=FALSE
# globalPara$errFlag=TRUE
# globalPara$powertext=""
# globalPara$boundary=c()
# globalPara$prior=c()
# }
#
# else{
#
# if(is.null(oc.mat.efftoxL) | length(oc.mat.efftoxL)==0){
# if(length(oc.mat.efftoxS)==4) optim <- oc.mat.efftoxS
# else optim=oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
#
# }else{
# if(length(oc.mat.efftoxS)==4) optimS <- oc.mat.efftoxS
# else optimS <- oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
# if(length(oc.mat.efftoxL)==4) optimL <- oc.mat.efftoxL
# else optimL <- oc.mat.efftoxL[oc.mat.efftoxL[,3]==min(oc.mat.efftoxL[,3]),]
#
# if(is.matrix(optimS)) optimS=optimS[1,]
# if(is.matrix(optimL)) optimL=optimL[1,]
#
# #message(optimS)
# #message(optimL)
#
# if(optimS[4]>=optimL[4]) {optim=optimS}
# else if(abs(optimS[3]-input$err1)>=abs(optimL[3]-input$err1) & (!input$truet1e)) {optim=optimL}
# else {optim=optimS}
# }
#
# if(is.matrix(optim)) optim=optim[1,]
globalPara$lambda_E <- optim[1]
globalPara$lambda_T <- optim[2]
globalPara$gamma <- optim[3]
globalPara$t1e <- optim[6]
# globalPara$power= optim[6]
globalPara$a01= optim[4]
globalPara$a10= optim[5]
globalPara$a00= optim[7]
cat("Prior is " ,prior)
boundary2=get_boundarycpp(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = optim[1], lambda_t =optim[2],
gamma = optim[3], r0=r0, t0=t0)
globalPara$boundary.eff = boundary2[[1]]
globalPara$boundary.tox = boundary2[[2]]
bound_temp=get_boundary_table(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = optim[1], lambda_t =optim[2],
gamma = optim[3], r0=r0, t0=t0)
globalPara$power = BOP2_TE.OC(interm.eff=nobs.seq,
interm.tox=nobs.seqTX,
boundary.eff=globalPara$boundary.eff,
boundary.tox=globalPara$boundary.tox,
scenario=scen[4,])[1]
# globalPara$boundary =cbind(unique(sort(c(interm.eff, interm.tox))),as.character(bound_temp[[1]]),
# as.character(bound_temp[[2]]))
globalPara$pts_e = optim[8]
globalPara$pts_t = optim[9]
globalPara$boundary =bound_temp
# dimnames(globalPara$boundary) <- list(NULL,c("# patients treated","Stop if # response <="," OR # toxicity >="))
globalPara$calculated=TRUE
# temp2 = Calc$Getoc_Rcpp_Tox_Eff(input$seed, input$numOfSimForTiralSetting, matrix(c(1,1,0,0,0,1,0,1),
# nrow=2,byrow=TRUE), nobs.seq,nobs.seqTX, optim[1], optim[2], prior, p.a, c(input$e1n,1-input$e2n), maxresp,resp_list_r);
# globalPara$power=temp2[[3]]
globalPara$powertext <- paste("The power of this trial is: ",round(globalPara$power,4),sep="")
isSBcalculated$ET <- TRUE
globalPara$prior=prior
}
}
else{globalPara$error="Error: The value of Prob(Eff & Tox) in Prior Specification is invalid (either too large or too small).
Given the values of Prob(Eff) and Prob(Tox), Prob(Eff & Tox) must satisfy the constraint Prob(Eff) + Prob(Tox) -1 <= Prob(Eff & Tox) <= min{Prob(Eff), Prob(Tox)}.
The left-hand side of the constraint is to ensure that Prob(no Eff & no Tox)>=0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()
}
}
else {globalPara$error="Error: The value in Prior Specification is missing or invalid (either too large or too small)."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of Pr(Eff & Tox) is invalid (either too large or too small).
Given the values of Pr(Eff) and Pr(Tox), Pr(Eff & Tox) must satisfy the constraint Pr(Eff) + Pr(Tox) -1 < Pr(Eff & Tox) < min{Pr(Eff), Pr(Tox)}.
The left-hand side of the constraint is to ensure that Pr(no Eff & no Tox)>0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
} else {globalPara$error="Error: The value of Pr(Eff)/Pr(Tox) in alternative hypothesis must be greater/smaller than the null."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of null/alternative hypothesis is missing, too small (negative) or greater than 1"
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
return(globalPara)
}
DoStoppingBoundaries_minN = function( input, globalPara, Calc, nobs.seq, nobs.seqTX, isSBcalculated, session )
{
#browser();
if(input$e1n>0.0001 & input$e1a>0.0001 & input$e1n<1 & input$e1a<1 &
input$e2n>0.0001 & input$e2a>0.0001 & input$e2n<1 & input$e2a<1 &
input$e3n>0.0001 & input$e3a>0.0001 & input$e3n<1 & input$e3a<1 &
!is.na(input$e1n) & !is.na(input$e1a) & !is.na(input$e2n) & !is.na(input$e2a) & !is.na(input$e3n) & !is.na(input$e3a)){
if(input$e1n<=input$e1a & input$e2n>=input$e2a) {
if(input$e1n+input$e2n-input$e3n<1 & input$e1a+input$e2a-input$e3a<1 &
input$e3n<min(input$e1n,input$e2n) & input$e3a<min(input$e1a,input$e2a)){
if((input$ETprior1>=0 & input$ETprior1<=1 & input$ETprior2>=0 & input$ETprior2<=1 & input$ETprior3>=0 & input$ETprior3<=1 &
!is.na(input$ETprior1) & !is.na(input$ETprior2) & !is.na(input$ETprior3) & !is.na(input$ETpriorSS) & input$ETpriorSS>0)| input$priorspec){
if(input$ETprior1+input$ETprior2-input$ETprior3 <= 1 & input$ETprior3 <= min(input$ETprior1,input$ETprior2)){
# cut.seq = sort(c(seq(0.5,0.95,by=0.025),seq(0.96,0.99,0.01),seq(0.9-input$err1,1-input$err1,0.005)))
cut.seq = rev(c(seq(0.5,0.8,0.025),seq(0.81,0.96,0.01)))
power.seq = log(seq(1,0.5,by=-0.025))/log(0.5)
p.n = c(input$e3n,input$e1n-input$e3n,input$e2n-input$e3n,1-input$e1n-input$e2n+input$e3n)
##input$e3n: Pr(Eff & Tox), input$e1n: Pr(Eff), input$e2n: Pr(Tox)
##input$e1n-input$e3n: Pr(Eff & no Tox)
##input$e2n-input$e3n: Pr(no Eff & Tox)
##1-input$e1n-input$e2n+input$e3n: Pr(no Eff & no Tox)
p.a = c(input$e3a,input$e1a-input$e3a,input$e2a-input$e3a,1-input$e1a-input$e2a+input$e3a)
# if(input$priorspec){
# prior = p.n
# }else{
# prior = c(input$ETprior3*input$ETpriorSS,(input$ETprior1-input$ETprior3)*input$ETpriorSS,
# (input$ETprior2-input$ETprior3)*input$ETpriorSS,(1-input$ETprior1-input$ETprior2+input$ETprior3)*input$ETpriorSS)
# }
oc.mat.efftox = c()
# oc.mat.efftox = Calc$GridSearchToxEffRcpp(seed=input$seed, contrast = matrix(c(1,1,0,0,0,1,0,1),nrow=2,byrow=TRUE),
# nobs = nobs.seq, nobsTX = nobs.seqTX, dprior = p.n,
# b = cut.seq, pow = power.seq, pn=p.n, pa = p.a,
# phi=c(input$e1n,1-input$e2n), cutstart=cut.start,
# nsim = input$numOfSimForTiralSetting,
# err1 = input$err1, maxresp,resp_list_r)
r0 = input$e1n
t0 = input$e2n
r1 = input$e1a
t1 = input$e2a
interm.eff = nobs.seq
interm.tox = nobs.seqTX
message(unique(sort(c(interm.eff, interm.tox))))
scen = hypotheses_ind(r0, r1, t0, t1)
#
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
# oc.mat.efftox.temp= grid_search_cpp_minN(interm.eff=interm.eff,
# interm.tox=interm.tox,
# scenario=scen,
# r0 = r0, r1= r1,
# t0=t0, t1= t1,
# lambda_r=cut.seq,
# lambda_t= cut.seq,
# gamma_space=power.seq,
# typeI01=input$err_eff, typeI10=input$err_tox,
# typeI00=input$err_all,
# power_e = input$power_eff,
# power_t = input$power_tox,
# true_opt = input$truet1e,
# prior = scen[1,],
# cppfunction=GridSearchBiRcpp) # return a vector
# oc.mat.efftox.temp= grid_search_cpp(interm.eff=interm.eff,
# interm.tox=interm.tox,
# scenario=scen,
# r0 = r0, r1= r1,
# t0=t0, t1= t1,
# lambda_r=cut.seq,
# lambda_t= cut.seq,
# gamma_space=power.seq,
# typeI01=input$err_eff, typeI10=input$err_tox,
# typeI00=input$err_all,
# true_opt = input$truet1e,
# prior = scen[1,],
# cppfunction=GridSearchBiRcpp) # return a vector
if(input$e3n == (input$e1n*input$e2n)& input$e3a == (input$e1a*input$e2a) ){
oc.mat.efftox =oc.mat.efftox.temp
message("Independent")
}else{
message("Not independent")
# optim.temp=oc.mat.efftox.temp[1:3]
# lambda_E_space = seq(optim.temp[1]-0.2, min(1,optim.temp[1]+0.05 ),by=0.01)
# lambda_T_space = seq(optim.temp[2]-0.2, min(1,optim.temp[2]+0.05 ),by=0.01)
# gamma_space = seq(optim.temp[3]-0.2, min(1,optim.temp[3]+0.1 ),by=0.02)
lambda_E_space = cut.seq
lambda_E2_space = cut.seq
lambda_T_space = cut.seq
lambda_T2_space = cut.seq
gamma_space = power.seq
scen = hypotheses_corr(r0,r1,t0,t1,PA_ET=input$e3a, PN_ET = input$e3n)
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
search_space = bond.unique(interm.eff=interm.eff, interm.tox=interm.tox, scenario=scen,
lambda_r = lambda_E_space,
lambda_t = lambda_T_space,
gamma_space= gamma_space,
r0=r0,t0=t0,
prior = scen[1,])
# search_space = bond.unique_2lambda(interm.eff=interm.eff, interm.tox=interm.tox, scenario=scen,
# lambda_r = lambda_E_space,
# lambda_t = lambda_T_space,
# lambda_r2 = lambda_E_space,
# lambda_t2 = lambda_T_space,
# gamma_space= gamma_space,
# r0=r0,t0=t0,
# prior = scen[1,])
oc.mat.efftox=optimization_corr_minN(interm.eff=interm.eff,
interm.tox=interm.tox,
search.temp = search_space,
scenario=scen,
r0=r0, t0=t0,
typeI01=input$err_eff,
typeI10=input$err_tox,
typeI00=input$err_all,
power11 = 0.8,
lpow=0.3,
ncore=1)
}
if(is.null(oc.mat.efftox) | length(oc.mat.efftox)==0) {
globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$boundary=c()
globalPara$prior=c()
} else {
optim=oc.mat.efftox
#message(oc.mat.efftox[1:10,])
# oc.mat.efftoxS = oc.mat.efftox[oc.mat.efftox[,3]<=input$err1,]
# oc.mat.efftoxL = oc.mat.efftox[oc.mat.efftox[,3]>input$err1,]
#
# #message(oc.mat.efftoxS)
# #message(oc.mat.efftoxL)
#
# if(is.null(oc.mat.efftoxS) | length(oc.mat.efftoxS)==0) {
# globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
# globalPara$calculated=FALSE
# globalPara$errFlag=TRUE
# globalPara$powertext=""
# globalPara$boundary=c()
# globalPara$prior=c()
# }
#
# else{
#
# if(is.null(oc.mat.efftoxL) | length(oc.mat.efftoxL)==0){
# if(length(oc.mat.efftoxS)==4) optim <- oc.mat.efftoxS
# else optim=oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
#
# }else{
# if(length(oc.mat.efftoxS)==4) optimS <- oc.mat.efftoxS
# else optimS <- oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
# if(length(oc.mat.efftoxL)==4) optimL <- oc.mat.efftoxL
# else optimL <- oc.mat.efftoxL[oc.mat.efftoxL[,3]==min(oc.mat.efftoxL[,3]),]
#
# if(is.matrix(optimS)) optimS=optimS[1,]
# if(is.matrix(optimL)) optimL=optimL[1,]
#
# #message(optimS)
# #message(optimL)
#
# if(optimS[4]>=optimL[4]) {optim=optimS}
# else if(abs(optimS[3]-input$err1)>=abs(optimL[3]-input$err1) & (!input$truet1e)) {optim=optimL}
# else {optim=optimS}
# }
#
# if(is.matrix(optim)) optim=optim[1,]
globalPara$lambda_E <- optim$parameter[1]
globalPara$lambda_T <- optim$parameter[2]
globalPara$gamma <- optim$parameter[3]
# globalPara$t1e <- optim$prob[6]
# globalPara$power= optim[6]
globalPara$a01= optim$prob["a01"]
globalPara$a10= optim$prob["a10"]
globalPara$a00= optim$prob["a00"]
globalPara$power = optim$prob["power"]
message("Prior is " ,prior)
boundary2=get_boundarycpp(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = globalPara$lambda_E, lambda_t = globalPara$lambda_T,
gamma = globalPara$gamma, r0=r0, t0=t0)
globalPara$boundary.eff = boundary2[[1]]
globalPara$boundary.tox = boundary2[[2]]
bound_temp=get_boundary_table(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = globalPara$lambda_E, lambda_t =globalPara$lambda_T,
gamma = globalPara$gamma, r0=r0, t0=t0)
# globalPara$power = BOP2_TE.OC(interm.eff=nobs.seq,
# interm.tox=nobs.seqTX,
# boundary.eff=globalPara$boundary.eff,
# boundary.tox=globalPara$boundary.tox,
# scenario=scen[4,])[1]
# globalPara$boundary =cbind(unique(sort(c(interm.eff, interm.tox))),as.character(bound_temp[[1]]),
# as.character(bound_temp[[2]]))
globalPara$samplesize = optim$SampleSize
globalPara$boundary =bound_temp
# dimnames(globalPara$boundary) <- list(NULL,c("# patients treated","Stop if # response <="," OR # toxicity >="))
globalPara$calculated=TRUE
# temp2 = Calc$Getoc_Rcpp_Tox_Eff(input$seed, input$numOfSimForTiralSetting, matrix(c(1,1,0,0,0,1,0,1),
# nrow=2,byrow=TRUE), nobs.seq,nobs.seqTX, optim[1], optim[2], prior, p.a, c(input$e1n,1-input$e2n), maxresp,resp_list_r);
# globalPara$power=temp2[[3]]
globalPara$powertext <- paste("The power of this trial is: ",round(globalPara$power,4),sep="")
isSBcalculated$ET <- TRUE
globalPara$prior=prior
}
}
else{globalPara$error="Error: The value of Prob(Eff & Tox) in Prior Specification is invalid (either too large or too small).
Given the values of Prob(Eff) and Prob(Tox), Prob(Eff & Tox) must satisfy the constraint Prob(Eff) + Prob(Tox) -1 <= Prob(Eff & Tox) <= min{Prob(Eff), Prob(Tox)}.
The left-hand side of the constraint is to ensure that Prob(no Eff & no Tox)>=0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()
}
}
else {globalPara$error="Error: The value in Prior Specification is missing or invalid (either too large or too small)."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of Pr(Eff & Tox) is invalid (either too large or too small).
Given the values of Pr(Eff) and Pr(Tox), Pr(Eff & Tox) must satisfy the constraint Pr(Eff) + Pr(Tox) -1 < Pr(Eff & Tox) < min{Pr(Eff), Pr(Tox)}.
The left-hand side of the constraint is to ensure that Pr(no Eff & no Tox)>0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
} else {globalPara$error="Error: The value of Pr(Eff)/Pr(Tox) in alternative hypothesis must be greater/smaller than the null."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of null/alternative hypothesis is missing, too small (negative) or greater than 1"
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
return(globalPara)
}
# end function DoStoppingBoundaries
Try the GBOP2 package in your browser
Any scripts or data that you put into this service are public.
GBOP2 documentation built on April 11, 2025, 5:42 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions DoStoppingBoundaries_minN DoStoppingBoundaries
ncore <- 1
# ------------------------
DoStoppingBoundaries = function( input, globalPara, Calc, nobs.seq, nobs.seqTX, isSBcalculated, session )
{
#browser();
if(input$e1n>0.0001 & input$e1a>0.0001 & input$e1n<1 & input$e1a<1 &
input$e2n>0.0001 & input$e2a>0.0001 & input$e2n<1 & input$e2a<1 &
input$e3n>0.0001 & input$e3a>0.0001 & input$e3n<1 & input$e3a<1 &
!is.na(input$e1n) & !is.na(input$e1a) & !is.na(input$e2n) & !is.na(input$e2a) & !is.na(input$e3n) & !is.na(input$e3a)){
if(input$e1n<=input$e1a & input$e2n>=input$e2a) {
if(input$e1n+input$e2n-input$e3n<1 & input$e1a+input$e2a-input$e3a<1 &
input$e3n<min(input$e1n,input$e2n) & input$e3a<min(input$e1a,input$e2a)){
if((input$ETprior1>=0 & input$ETprior1<=1 & input$ETprior2>=0 & input$ETprior2<=1 & input$ETprior3>=0 & input$ETprior3<=1 &
!is.na(input$ETprior1) & !is.na(input$ETprior2) & !is.na(input$ETprior3) & !is.na(input$ETpriorSS) & input$ETpriorSS>0)| input$priorspec){
if(input$ETprior1+input$ETprior2-input$ETprior3 <= 1 & input$ETprior3 <= min(input$ETprior1,input$ETprior2)){
# cut.seq = sort(c(seq(0.5,0.95,by=0.025),seq(0.96,0.99,0.01),seq(0.9-input$err1,1-input$err1,0.005)))
cut.seq = rev(c(seq(0.5,0.8,0.025),seq(0.81,0.96,0.01)))
power.seq = log(seq(1,0.5,by=-0.025))/log(0.5)
p.n = c(input$e3n,input$e1n-input$e3n,input$e2n-input$e3n,1-input$e1n-input$e2n+input$e3n)
##input$e3n: Pr(Eff & Tox), input$e1n: Pr(Eff), input$e2n: Pr(Tox)
##input$e1n-input$e3n: Pr(Eff & no Tox)
##input$e2n-input$e3n: Pr(no Eff & Tox)
##1-input$e1n-input$e2n+input$e3n: Pr(no Eff & no Tox)
p.a = c(input$e3a,input$e1a-input$e3a,input$e2a-input$e3a,1-input$e1a-input$e2a+input$e3a)
# if(input$priorspec){
# prior = p.n
# }else{
# prior = c(input$ETprior3*input$ETpriorSS,(input$ETprior1-input$ETprior3)*input$ETpriorSS,
# (input$ETprior2-input$ETprior3)*input$ETpriorSS,(1-input$ETprior1-input$ETprior2+input$ETprior3)*input$ETpriorSS)
# }
oc.mat.efftox = c()
cut.start = 1
# oc.mat.efftox = Calc$GridSearchToxEffRcpp(seed=input$seed, contrast = matrix(c(1,1,0,0,0,1,0,1),nrow=2,byrow=TRUE),
# nobs = nobs.seq, nobsTX = nobs.seqTX, dprior = p.n,
# b = cut.seq, pow = power.seq, pn=p.n, pa = p.a,
# phi=c(input$e1n,1-input$e2n), cutstart=cut.start,
# nsim = input$numOfSimForTiralSetting,
# err1 = input$err1, maxresp,resp_list_r)
########################################################
r0 = input$e1n
t0 = input$e2n
r1 = input$e1a
t1 = input$e2a
interm.eff = nobs.seq #
interm.tox = nobs.seqTX #
message(unique(sort(c(interm.eff, interm.tox)))) #
scen = hypotheses_ind(r0, r1, t0, t1) ##independent
##########################################################
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
oc.mat.efftox.temp= grid_search_cpp(interm.eff=interm.eff,
interm.tox=interm.tox,
scenario=scen,
r0 = r0, r1= r1,
t0=t0, t1= t1,
lambda_r=cut.seq,
lambda_t= cut.seq,
gamma_space=power.seq,
typeI01=input$err_eff, typeI10=input$err_tox,
typeI00=input$err_all,
true_opt = input$truet1e,
prior = scen[1,],
cppfunction=GridSearchBiRcpp) # return a vector
if(input$e3n == (input$e1n*input$e2n)& input$e3a == (input$e1a*input$e2a) ){
oc.mat.efftox =oc.mat.efftox.temp
message("Independent")
}else{
message("Not independent")
optim.temp=oc.mat.efftox.temp[1:3]
lambda_E_space = seq(optim.temp[1]-0.2, min(1,optim.temp[1]+0.05 ),by=0.01)
lambda_T_space = seq(optim.temp[2]-0.2, min(1,optim.temp[2]+0.05 ),by=0.01)
gamma_space = seq(optim.temp[3]-0.2, min(1,optim.temp[3]+0.1 ),by=0.02)
scen = hypotheses_corr(r0,r1,t0,t1,PA_ET=input$e3a, PN_ET = input$e3n)
message(scen)
# joint probabilities of four possible outcomes
# (no efficacy and no toxicity, no efficacy and toxicity, efficacy and no toxicity, both efficacy and toxicity)
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
search_space = bond.unique(interm.eff=interm.eff, interm.tox=interm.tox, scenario=scen,
lambda_r = lambda_E_space,
lambda_t = lambda_T_space,
gamma_space= gamma_space,
r0=r0,t0=t0,
prior = scen[1,])
oc.mat.efftox=optimization_corr(interm.eff=interm.eff,
interm.tox=interm.tox,
search.temp = search_space,
scenario=scen,
r0=r0, t0=t0,
typeI01=input$err_eff,
typeI10=input$err_tox,
typeI00=input$err_all,
lpow=0.3,
ncore=1)
}
if(is.null(oc.mat.efftox) | length(oc.mat.efftox)==0) {
globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$boundary=c()
globalPara$prior=c()
} else {
optim=oc.mat.efftox
#message(oc.mat.efftox[1:10,])
# oc.mat.efftoxS = oc.mat.efftox[oc.mat.efftox[,3]<=input$err1,]
# oc.mat.efftoxL = oc.mat.efftox[oc.mat.efftox[,3]>input$err1,]
#
# #message(oc.mat.efftoxS)
# #message(oc.mat.efftoxL)
#
# if(is.null(oc.mat.efftoxS) | length(oc.mat.efftoxS)==0) {
# globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
# globalPara$calculated=FALSE
# globalPara$errFlag=TRUE
# globalPara$powertext=""
# globalPara$boundary=c()
# globalPara$prior=c()
# }
#
# else{
#
# if(is.null(oc.mat.efftoxL) | length(oc.mat.efftoxL)==0){
# if(length(oc.mat.efftoxS)==4) optim <- oc.mat.efftoxS
# else optim=oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
#
# }else{
# if(length(oc.mat.efftoxS)==4) optimS <- oc.mat.efftoxS
# else optimS <- oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
# if(length(oc.mat.efftoxL)==4) optimL <- oc.mat.efftoxL
# else optimL <- oc.mat.efftoxL[oc.mat.efftoxL[,3]==min(oc.mat.efftoxL[,3]),]
#
# if(is.matrix(optimS)) optimS=optimS[1,]
# if(is.matrix(optimL)) optimL=optimL[1,]
#
# #message(optimS)
# #message(optimL)
#
# if(optimS[4]>=optimL[4]) {optim=optimS}
# else if(abs(optimS[3]-input$err1)>=abs(optimL[3]-input$err1) & (!input$truet1e)) {optim=optimL}
# else {optim=optimS}
# }
#
# if(is.matrix(optim)) optim=optim[1,]
globalPara$lambda_E <- optim[1]
globalPara$lambda_T <- optim[2]
globalPara$gamma <- optim[3]
globalPara$t1e <- optim[6]
# globalPara$power= optim[6]
globalPara$a01= optim[4]
globalPara$a10= optim[5]
globalPara$a00= optim[7]
cat("Prior is " ,prior)
boundary2=get_boundarycpp(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = optim[1], lambda_t =optim[2],
gamma = optim[3], r0=r0, t0=t0)
globalPara$boundary.eff = boundary2[[1]]
globalPara$boundary.tox = boundary2[[2]]
bound_temp=get_boundary_table(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = optim[1], lambda_t =optim[2],
gamma = optim[3], r0=r0, t0=t0)
globalPara$power = BOP2_TE.OC(interm.eff=nobs.seq,
interm.tox=nobs.seqTX,
boundary.eff=globalPara$boundary.eff,
boundary.tox=globalPara$boundary.tox,
scenario=scen[4,])[1]
# globalPara$boundary =cbind(unique(sort(c(interm.eff, interm.tox))),as.character(bound_temp[[1]]),
# as.character(bound_temp[[2]]))
globalPara$pts_e = optim[8]
globalPara$pts_t = optim[9]
globalPara$boundary =bound_temp
# dimnames(globalPara$boundary) <- list(NULL,c("# patients treated","Stop if # response <="," OR # toxicity >="))
globalPara$calculated=TRUE
# temp2 = Calc$Getoc_Rcpp_Tox_Eff(input$seed, input$numOfSimForTiralSetting, matrix(c(1,1,0,0,0,1,0,1),
# nrow=2,byrow=TRUE), nobs.seq,nobs.seqTX, optim[1], optim[2], prior, p.a, c(input$e1n,1-input$e2n), maxresp,resp_list_r);
# globalPara$power=temp2[[3]]
globalPara$powertext <- paste("The power of this trial is: ",round(globalPara$power,4),sep="")
isSBcalculated$ET <- TRUE
globalPara$prior=prior
}
}
else{globalPara$error="Error: The value of Prob(Eff & Tox) in Prior Specification is invalid (either too large or too small).
Given the values of Prob(Eff) and Prob(Tox), Prob(Eff & Tox) must satisfy the constraint Prob(Eff) + Prob(Tox) -1 <= Prob(Eff & Tox) <= min{Prob(Eff), Prob(Tox)}.
The left-hand side of the constraint is to ensure that Prob(no Eff & no Tox)>=0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()
}
}
else {globalPara$error="Error: The value in Prior Specification is missing or invalid (either too large or too small)."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of Pr(Eff & Tox) is invalid (either too large or too small).
Given the values of Pr(Eff) and Pr(Tox), Pr(Eff & Tox) must satisfy the constraint Pr(Eff) + Pr(Tox) -1 < Pr(Eff & Tox) < min{Pr(Eff), Pr(Tox)}.
The left-hand side of the constraint is to ensure that Pr(no Eff & no Tox)>0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
} else {globalPara$error="Error: The value of Pr(Eff)/Pr(Tox) in alternative hypothesis must be greater/smaller than the null."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of null/alternative hypothesis is missing, too small (negative) or greater than 1"
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
return(globalPara)
}
DoStoppingBoundaries_minN = function( input, globalPara, Calc, nobs.seq, nobs.seqTX, isSBcalculated, session )
{
#browser();
if(input$e1n>0.0001 & input$e1a>0.0001 & input$e1n<1 & input$e1a<1 &
input$e2n>0.0001 & input$e2a>0.0001 & input$e2n<1 & input$e2a<1 &
input$e3n>0.0001 & input$e3a>0.0001 & input$e3n<1 & input$e3a<1 &
!is.na(input$e1n) & !is.na(input$e1a) & !is.na(input$e2n) & !is.na(input$e2a) & !is.na(input$e3n) & !is.na(input$e3a)){
if(input$e1n<=input$e1a & input$e2n>=input$e2a) {
if(input$e1n+input$e2n-input$e3n<1 & input$e1a+input$e2a-input$e3a<1 &
input$e3n<min(input$e1n,input$e2n) & input$e3a<min(input$e1a,input$e2a)){
if((input$ETprior1>=0 & input$ETprior1<=1 & input$ETprior2>=0 & input$ETprior2<=1 & input$ETprior3>=0 & input$ETprior3<=1 &
!is.na(input$ETprior1) & !is.na(input$ETprior2) & !is.na(input$ETprior3) & !is.na(input$ETpriorSS) & input$ETpriorSS>0)| input$priorspec){
if(input$ETprior1+input$ETprior2-input$ETprior3 <= 1 & input$ETprior3 <= min(input$ETprior1,input$ETprior2)){
# cut.seq = sort(c(seq(0.5,0.95,by=0.025),seq(0.96,0.99,0.01),seq(0.9-input$err1,1-input$err1,0.005)))
cut.seq = rev(c(seq(0.5,0.8,0.025),seq(0.81,0.96,0.01)))
power.seq = log(seq(1,0.5,by=-0.025))/log(0.5)
p.n = c(input$e3n,input$e1n-input$e3n,input$e2n-input$e3n,1-input$e1n-input$e2n+input$e3n)
##input$e3n: Pr(Eff & Tox), input$e1n: Pr(Eff), input$e2n: Pr(Tox)
##input$e1n-input$e3n: Pr(Eff & no Tox)
##input$e2n-input$e3n: Pr(no Eff & Tox)
##1-input$e1n-input$e2n+input$e3n: Pr(no Eff & no Tox)
p.a = c(input$e3a,input$e1a-input$e3a,input$e2a-input$e3a,1-input$e1a-input$e2a+input$e3a)
# if(input$priorspec){
# prior = p.n
# }else{
# prior = c(input$ETprior3*input$ETpriorSS,(input$ETprior1-input$ETprior3)*input$ETpriorSS,
# (input$ETprior2-input$ETprior3)*input$ETpriorSS,(1-input$ETprior1-input$ETprior2+input$ETprior3)*input$ETpriorSS)
# }
oc.mat.efftox = c()
# oc.mat.efftox = Calc$GridSearchToxEffRcpp(seed=input$seed, contrast = matrix(c(1,1,0,0,0,1,0,1),nrow=2,byrow=TRUE),
# nobs = nobs.seq, nobsTX = nobs.seqTX, dprior = p.n,
# b = cut.seq, pow = power.seq, pn=p.n, pa = p.a,
# phi=c(input$e1n,1-input$e2n), cutstart=cut.start,
# nsim = input$numOfSimForTiralSetting,
# err1 = input$err1, maxresp,resp_list_r)
r0 = input$e1n
t0 = input$e2n
r1 = input$e1a
t1 = input$e2a
interm.eff = nobs.seq
interm.tox = nobs.seqTX
message(unique(sort(c(interm.eff, interm.tox))))
scen = hypotheses_ind(r0, r1, t0, t1)
#
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
# oc.mat.efftox.temp= grid_search_cpp_minN(interm.eff=interm.eff,
# interm.tox=interm.tox,
# scenario=scen,
# r0 = r0, r1= r1,
# t0=t0, t1= t1,
# lambda_r=cut.seq,
# lambda_t= cut.seq,
# gamma_space=power.seq,
# typeI01=input$err_eff, typeI10=input$err_tox,
# typeI00=input$err_all,
# power_e = input$power_eff,
# power_t = input$power_tox,
# true_opt = input$truet1e,
# prior = scen[1,],
# cppfunction=GridSearchBiRcpp) # return a vector
# oc.mat.efftox.temp= grid_search_cpp(interm.eff=interm.eff,
# interm.tox=interm.tox,
# scenario=scen,
# r0 = r0, r1= r1,
# t0=t0, t1= t1,
# lambda_r=cut.seq,
# lambda_t= cut.seq,
# gamma_space=power.seq,
# typeI01=input$err_eff, typeI10=input$err_tox,
# typeI00=input$err_all,
# true_opt = input$truet1e,
# prior = scen[1,],
# cppfunction=GridSearchBiRcpp) # return a vector
if(input$e3n == (input$e1n*input$e2n)& input$e3a == (input$e1a*input$e2a) ){
oc.mat.efftox =oc.mat.efftox.temp
message("Independent")
}else{
message("Not independent")
# optim.temp=oc.mat.efftox.temp[1:3]
# lambda_E_space = seq(optim.temp[1]-0.2, min(1,optim.temp[1]+0.05 ),by=0.01)
# lambda_T_space = seq(optim.temp[2]-0.2, min(1,optim.temp[2]+0.05 ),by=0.01)
# gamma_space = seq(optim.temp[3]-0.2, min(1,optim.temp[3]+0.1 ),by=0.02)
lambda_E_space = cut.seq
lambda_E2_space = cut.seq
lambda_T_space = cut.seq
lambda_T2_space = cut.seq
gamma_space = power.seq
scen = hypotheses_corr(r0,r1,t0,t1,PA_ET=input$e3a, PN_ET = input$e3n)
if(input$priorspec){
prior = scen[1,]
}else{
# pi = c(p00, p01, p10, p11)
prior= c(1-input$ETprior1-input$ETprior2+input$ETprior3,
input$ETprior2-input$ETprior3,
input$ETprior1-input$ETprior3,
input$ETprior3)*input$ETpriorSS
}
search_space = bond.unique(interm.eff=interm.eff, interm.tox=interm.tox, scenario=scen,
lambda_r = lambda_E_space,
lambda_t = lambda_T_space,
gamma_space= gamma_space,
r0=r0,t0=t0,
prior = scen[1,])
# search_space = bond.unique_2lambda(interm.eff=interm.eff, interm.tox=interm.tox, scenario=scen,
# lambda_r = lambda_E_space,
# lambda_t = lambda_T_space,
# lambda_r2 = lambda_E_space,
# lambda_t2 = lambda_T_space,
# gamma_space= gamma_space,
# r0=r0,t0=t0,
# prior = scen[1,])
oc.mat.efftox=optimization_corr_minN(interm.eff=interm.eff,
interm.tox=interm.tox,
search.temp = search_space,
scenario=scen,
r0=r0, t0=t0,
typeI01=input$err_eff,
typeI10=input$err_tox,
typeI00=input$err_all,
power11 = 0.8,
lpow=0.3,
ncore=1)
}
if(is.null(oc.mat.efftox) | length(oc.mat.efftox)==0) {
globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$boundary=c()
globalPara$prior=c()
} else {
optim=oc.mat.efftox
#message(oc.mat.efftox[1:10,])
# oc.mat.efftoxS = oc.mat.efftox[oc.mat.efftox[,3]<=input$err1,]
# oc.mat.efftoxL = oc.mat.efftox[oc.mat.efftox[,3]>input$err1,]
#
# #message(oc.mat.efftoxS)
# #message(oc.mat.efftoxL)
#
# if(is.null(oc.mat.efftoxS) | length(oc.mat.efftoxS)==0) {
# globalPara$error="The power is too small (<5%) under the current setting. Please increase the type I error or the sample size."
# globalPara$calculated=FALSE
# globalPara$errFlag=TRUE
# globalPara$powertext=""
# globalPara$boundary=c()
# globalPara$prior=c()
# }
#
# else{
#
# if(is.null(oc.mat.efftoxL) | length(oc.mat.efftoxL)==0){
# if(length(oc.mat.efftoxS)==4) optim <- oc.mat.efftoxS
# else optim=oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
#
# }else{
# if(length(oc.mat.efftoxS)==4) optimS <- oc.mat.efftoxS
# else optimS <- oc.mat.efftoxS[oc.mat.efftoxS[,4]==max(oc.mat.efftoxS[,4]),]
# if(length(oc.mat.efftoxL)==4) optimL <- oc.mat.efftoxL
# else optimL <- oc.mat.efftoxL[oc.mat.efftoxL[,3]==min(oc.mat.efftoxL[,3]),]
#
# if(is.matrix(optimS)) optimS=optimS[1,]
# if(is.matrix(optimL)) optimL=optimL[1,]
#
# #message(optimS)
# #message(optimL)
#
# if(optimS[4]>=optimL[4]) {optim=optimS}
# else if(abs(optimS[3]-input$err1)>=abs(optimL[3]-input$err1) & (!input$truet1e)) {optim=optimL}
# else {optim=optimS}
# }
#
# if(is.matrix(optim)) optim=optim[1,]
globalPara$lambda_E <- optim$parameter[1]
globalPara$lambda_T <- optim$parameter[2]
globalPara$gamma <- optim$parameter[3]
# globalPara$t1e <- optim$prob[6]
# globalPara$power= optim[6]
globalPara$a01= optim$prob["a01"]
globalPara$a10= optim$prob["a10"]
globalPara$a00= optim$prob["a00"]
globalPara$power = optim$prob["power"]
message("Prior is " ,prior)
boundary2=get_boundarycpp(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = globalPara$lambda_E, lambda_t = globalPara$lambda_T,
gamma = globalPara$gamma, r0=r0, t0=t0)
globalPara$boundary.eff = boundary2[[1]]
globalPara$boundary.tox = boundary2[[2]]
bound_temp=get_boundary_table(interm.eff=interm.eff,
interm.tox=interm.tox,
prior=prior,
lambda_e = globalPara$lambda_E, lambda_t =globalPara$lambda_T,
gamma = globalPara$gamma, r0=r0, t0=t0)
# globalPara$power = BOP2_TE.OC(interm.eff=nobs.seq,
# interm.tox=nobs.seqTX,
# boundary.eff=globalPara$boundary.eff,
# boundary.tox=globalPara$boundary.tox,
# scenario=scen[4,])[1]
# globalPara$boundary =cbind(unique(sort(c(interm.eff, interm.tox))),as.character(bound_temp[[1]]),
# as.character(bound_temp[[2]]))
globalPara$samplesize = optim$SampleSize
globalPara$boundary =bound_temp
# dimnames(globalPara$boundary) <- list(NULL,c("# patients treated","Stop if # response <="," OR # toxicity >="))
globalPara$calculated=TRUE
# temp2 = Calc$Getoc_Rcpp_Tox_Eff(input$seed, input$numOfSimForTiralSetting, matrix(c(1,1,0,0,0,1,0,1),
# nrow=2,byrow=TRUE), nobs.seq,nobs.seqTX, optim[1], optim[2], prior, p.a, c(input$e1n,1-input$e2n), maxresp,resp_list_r);
# globalPara$power=temp2[[3]]
globalPara$powertext <- paste("The power of this trial is: ",round(globalPara$power,4),sep="")
isSBcalculated$ET <- TRUE
globalPara$prior=prior
}
}
else{globalPara$error="Error: The value of Prob(Eff & Tox) in Prior Specification is invalid (either too large or too small).
Given the values of Prob(Eff) and Prob(Tox), Prob(Eff & Tox) must satisfy the constraint Prob(Eff) + Prob(Tox) -1 <= Prob(Eff & Tox) <= min{Prob(Eff), Prob(Tox)}.
The left-hand side of the constraint is to ensure that Prob(no Eff & no Tox)>=0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()
}
}
else {globalPara$error="Error: The value in Prior Specification is missing or invalid (either too large or too small)."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of Pr(Eff & Tox) is invalid (either too large or too small).
Given the values of Pr(Eff) and Pr(Tox), Pr(Eff & Tox) must satisfy the constraint Pr(Eff) + Pr(Tox) -1 < Pr(Eff & Tox) < min{Pr(Eff), Pr(Tox)}.
The left-hand side of the constraint is to ensure that Pr(no Eff & no Tox)>0."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
} else {globalPara$error="Error: The value of Pr(Eff)/Pr(Tox) in alternative hypothesis must be greater/smaller than the null."
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
}
else {globalPara$error="Error: The value of null/alternative hypothesis is missing, too small (negative) or greater than 1"
globalPara$calculated=FALSE
globalPara$errFlag=TRUE
globalPara$powertext=""
globalPara$prior=c()
globalPara$boundary=c()}
return(globalPara)
}
# end function DoStoppingBoundaries
Try the GBOP2 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.