NextBestMaxGainSamples-class: Next best dose with maximum gain value based on a pseudo DLE...

In crmPack: Object-Oriented Implementation of CRM Designs

Next best dose with maximum gain value based on a pseudo DLE and efficacy model with samples

Description

This is a class for which to find the next dose which is safe and give the maximum gain value for allocation. This is a class where DLE and efficacy samples are involved. There are two inputs which are the two target probabilities of the occurrence of a DLE used during trial and used at the end of trial, for finding the next best dose that is safe and gives the maximum gain value and the dose to recommend at the end of a trial. This is only suitable to use with DLE models specified in 'ModelTox' class and efficacy models specified in 'ModelEff' class class

Slots

DLEDuringTrialtarget

the target probability of the occurrrence of a DLE to be used during the trial

DLEEndOfTrialtarget

the target probability of the occurrence of a DLE to be used at the end of the trial. This target is particularly used to recommend the dose for which its posterior probability of the occurrence of a DLE is equal to this target

TDderive

the function which derives from the input, a vector of the posterior samples called TDsamples of the dose which has the probability of the occurrence of DLE equals to either the targetDuringTrial or targetEndOfTrial, the final next best TDtargetDuringTrial (the dose with probability of the occurrence of DLE equals to the targetDuringTrial)and TDtargetEndOfTrial estimate.

Gstarderive

the function which derives from the input, a vector of the posterior Gstar (the dose which gives the maximum gain value) samples called Gstarsamples, the final next best Gstar estimate.

Examples

##define the NextBestMaxGainsamples class
##specified the target probability of the occurrence of a DLE during the trial be 0.35
##specified the target probability of the occurrence of a DLE at the end of trial be 0.3
## we want the use the 30% posterior quantile (the 30th percentaile) of the TD35 
## (the dose level with probability of DLE equals 0.35) and TD30 samples
## For Gstar (the dose which gives tha maximum
##gain) samples, we will use the 50% posterio quantile (the median or th 50th percentile) 
## of the Gstar sample
##A function is then defined in the 'TDderive' slot for the TD30 and TD35 samples
## and another function is defined in the 'Gstarderive' slot for Gstar samples
mynextbest<-NextBestMaxGainSamples(DLEDuringTrialtarget=0.35,
                                   DLEEndOfTrialtarget=0.3,
                                   TDderive=function(TDsamples){
                                     quantile(TDsamples,prob=0.3)},
                                   Gstarderive=function(Gstarsamples){
                                     quantile(Gstarsamples,prob=0.5)})

crmPack documentation built on June 26, 2024, 5:07 p.m.