unc_analysis_assessment_bp: Assessment model - Bounded Probability

In Iraices/PrecisePvsBoundedP: A suggestion for the quantification of precise and bounded probability to express epistemic uncertainty in scientific assessments

Description

This function does the aluminium exposure assessment. It estimates the expected value and the highest posterior density of the frequency of exceeding the threshold. Note: the difference respect to the precise probability is that a normal distribution is fitted to the elicited values from experts using two percentiles and by specifying bounds on a percentile.

Usage

unc_analysis_assessment_bp(
  niter_ale,
  niter_epi,
  threshold,
  exposure_scenario,
  suff_stat_concentration,
  suff_stat_consumption,
  consumption_change_vals_EKE,
  consumption_change_probs_EKE,
  consumers_info_sample_size,
  concentration_mu0,
  concentration_v0,
  concentration_alpha0,
  concentration_beta0,
  sufficient_statistics_concentration,
  consumption_mu0,
  consumption_v0,
  consumption_alpha0,
  consumption_beta0,
  sufficient_statistics_consumption,
  consumption_event_alpha0,
  consumption_event_beta0
)

Arguments

niter_ale	number of generated samples
niter_epi	number of generated parameters from the posterior distrbutions (it indicates the number of repetitions the assessment will be done)
threshold	safety threshold
exposure_scenario	a value that indicates if the assessment is done an average consumption scenario by 'av' or on a high consumption scenario by 'perc_95'. Default is NULL
suff_stat_concentration	a vector of sufficient statistics: sample_size, sample_mean and sample_sd corresponding to concentration. If sufficient_statistics_concentration = FALSE, then it is vector of observed data
suff_stat_consumption	a vector of sufficient statistics: sample_size, sample_mean and sample_sd then it is vector of observed data
consumption_change_vals_EKE	a vector of elicited values from experts
consumption_change_probs_EKE	a vector of elicited probabilities from experts
consumers_info_sample_size	a vector with the sample size of non_consumer_sample_size and consumer_sample_size
concentration_mu0	prior hyperparameter mu0 for the normal-gamma distribution corresponding to concentration
concentration_v0	prior hyperparameter v0 for the normal-gamma distribution corresponding to concentration
concentration_alpha0	prior hyperparameter alpha0 for the normal-gamma distribution corresponding to concentration
concentration_beta0	prior hyperparameter beta0 for the normal-gamma distribution corresponding to concentration
sufficient_statistics_concentration	logical; if TRUE, sufficient statistics (sample_size, sample_mean, sample_variance) corresponding to concentration are given as input data, otherwise sufficient_statistics_concentration is given as observed data. Default is TRUE
consumption_mu0	prior hyperparameter mu0 for the normal-gamma distribution corresponding to consumption
consumption_v0	prior hyperparameter v0 for the normal-gamma distribution corresponding to consumption
consumption_alpha0	prior hyperparameter alpha0 for the normal-gamma distribution corresponding to consumption
consumption_beta0	prior hyperparameter beta0 for the normal-gamma distribution corresponding to consumption
sufficient_statistics_consumption	logical; if TRUE, sufficient statistics (sample_size, sample_mean, sample_variance) corresponding to consumption are given as input data, otherwise sufficient_statistics_consumption is given as observed data. Default is TRUE
consumption_event_alpha0	prior hyperparameter alpha0 for the beta distribution corresponding to consumption event
consumption_event_beta0	prior hyperparameter beta0 for the beta distribution corresponding to consumption event

Value

a list with the following components

prob_consumption_event

The estimated probability of consumption events

parameters_concentration

A list with the values of the prior and posterior parameters of concentration

parameters_consumption

A list with the prior and posterior parameters of consumption

frequency_exceeding

A vector with the estimated frequency of exceeding the threshold (the lenght is niter_epi)

expected_frequency_exceeding

The expected value of the frequency of exceeding the threshold

hdi_frequency_exceeding

The highest posterior density interval of the frequency of exceeding the threshold

Iraices/PrecisePvsBoundedP documentation built on Jan. 18, 2021, 11:32 p.m.