An R6 class encapsulating Cumulative-Cohort Designs
An R6 class encapsulating Cumulative-Cohort Designs
precautionary::Cpe -> Ccd
new()Ccd$new(escalate, deescalate, eliminate, cohort_max, enroll_max)
escalateEscalation boundary
deescalateDeescalation boundary
eliminateElimination boundary
cohort_maxUpper bound on dose-wise enrollment
enroll_maxUpper bound on total enrollment
Create a new Ccd object.
A Ccd object.
# TODO
applied()Ccd$applied(x, o, last_dose, max_dose, ...)
xA dose-wise vector of toxicity counts
oA dose-wise vector of non-toxicity counts
last_doseThe most recently given dose, as required to implement cumulative-cohort-based escalation decisions.
max_doseAn upper limit on future dose levels
...Unused by Ccd; included for superclass method compatibility
Return dose recommendation for given tox/no-tox tallies.
An object with components:
$stop - logical value indicating whether stop is indicated
$mtd - integer value, the recommended dose
$max_dose - integer value, a dose not to be exceeded henceforth.
clone()The objects of this class are cloneable with this method.
Ccd$clone(deep = FALSE)
deepWhether to make a deep clone.
TODO: Explain the hierarchy of model classes, including connections
with the executable specifications set forth in exec/prolog/ccd.pl.
Ivanova A, Flournoy N, Chung Y. Cumulative cohort design for dose-finding. Journal of Statistical Planning and Inference. 2007;137(7):2316-2327. 10.1016/j.jspi.2006.07.009
Liu S, Yuan Y. Bayesian optimal interval designs for phase I clinical trials. J R Stat Soc C. 2015;64(3):507-523. 10.1111/rssc.12089
# NOT RUN {
## ------------------------------------------------
## Method `Ccd$new`
## ------------------------------------------------
# TODO
# }
Run the code above in your browser using DataLab