This is the model class for single agent dose escalation,
from which all other specific models inherit. It inherits all slots
from GeneralModel.
dosea function computing the dose reaching a specific target probability, based on the model parameters and additional prior settings (see the details above)
proba function computing the probability of toxicity for a specific dose, based on the model parameters and additional prior settings (see the details above)
The datamodel must obey the convention that the data input is
called exactly as in the Data class.
All prior distributions for parameters should be contained in the
model function priormodel. The background is that this can
be used to simulate from the prior distribution, before obtaining any
data.
The dose function has as first argument prob, a scalar
toxicity probability which is targeted. Additional arguments are model
parameters. Then it computes, using model parameter(s) (samples), the
resulting dose. Note that the model parameters are called exactly as in the
model and must be included in the sample vector. The vectors
of all samples for these parameters will then be supplied to the function.
So your function must be able to process vectors of the model parameters,
i.e. it must vectorize over them.
The prob function has as first argument dose, which is a
scalar dose. Additional arguments are model parameters. Then it computes,
using model parameter(s) (samples), the resulting probability of toxicity at
that dose. Again here, the function must vectorize over the model
parameters.
If you work with multivariate parameters, then please assume that your the two functions receive either one parameter value as a row vector, or a samples matrix where the rows correspond to the sampling index, i.e. the layout is then nSamples x dimParameter.
Note that dose and prob are the inverse functions of each
other.
LogisticNormal,
LogisticLogNormal,
LogisticLogNormalSub,
LogisticKadane,
DualEndpoint