This class pools the source and target studies before analysing
them with a uniform prior on each arm response rate. The posterior is
available in closed form, so it inherits from the BinomialConjugate class
rather than sampling.
Super classes
Model -> BinomialConjugate -> BinomialPooling
Methods
Inherited methods
Model$calibrate_for_design()Model$check_data()Model$create()Model$empirical_bayes_update()Model$estimate_bayesian_operating_characteristics()Model$estimate_frequentist_operating_characteristics()Model$hypothesis_space_transformation()Model$inference()Model$inference_cache_scope()Model$plot_pdfs()Model$plot_posterior_pdf()Model$plot_prior_pdf()Model$posterior_beta_mixture()Model$posterior_mean()Model$posterior_to_RBesT()Model$print_model_summary()Model$prior_elir_ess()Model$prior_to_RBesT()Model$simulation_for_given_treatment_effect()Model$test_decision()Model$vectorised_replicate_inference()BinomialConjugate$credible_interval()BinomialConjugate$initialize()BinomialConjugate$posterior_cdf()BinomialConjugate$posterior_ess()BinomialConjugate$posterior_median()BinomialConjugate$posterior_moments()BinomialConjugate$posterior_pdf()BinomialConjugate$posterior_quantile()BinomialConjugate$prior_cdf()BinomialConjugate$prior_pdf()BinomialConjugate$sample_posterior()BinomialConjugate$sample_prior()BinomialConjugate$summary_rows()
BinomialPooling$prior_given_control_rate()
The prior of the treatment effect given the target control rate
Pooling treats the source study's patients as the target's, so before any target patient is seen the response rates follow the source posterior under uniform priors, the two arms independently. Given the control rate, the treatment effect is the treatment rate less that rate, the treatment rate following its source posterior. The inherited marginal prior, a uniform prior on each arm, is the one pooling updates rather than the one its analysis assumes about the target.