Package index • vira

Problem setup

Define decision problems and display results.

voi_problem(): Define a Value of Information problem
voi_problem_from_dist(): Define a VOI problem from probability distributions
decision_table(): Build a formatted decision table from a VOI problem

Construct, elicit, and use risk preference objects.

risk_preference(): Create a risk preference object directly from known parameters
use_risk_preference(): Generate utility and inverse-utility closures from a risk preference object
elicit_risk_preferences(): Elicit risk preferences through interactive questions
plot(<risk_preference>): Plot diagnostics for a risk_preference object

Step-by-step pipeline for computing VOI. Each function advances the voi_problem object through one stage of the calculation.

transform_to_utility(): Transform action-state values to utility units
optimize_action(): Find the optimal action under certainty and uncertainty
calculate_utility_dist(): Evaluate the utility distribution across states for each experiment outcome
summarize_utility(): Summarise the utility distribution to a single number
transform_to_values(): Convert summarised utility back to original outcome units
calculate_value_info(): Calculate the Value of Information

Convenience wrappers that run the full pipeline in one call.

calculate_evpi(): Calculate EVPI in one step
calculate_evsi(): Calculate EVSI in one step
survey_worth_it(): Compare a proposed survey cost against the EVPI or EVSI
voi_sensitivity(): Compute VOI across a range of risk aversion parameter values

Scalar summaries of outcome distributions (used inside the pipeline).

EV_voi(): Expected value (probability-weighted mean)
ES_voi(): Expected shortfall (Conditional Value-at-Risk)
max_voi(): Maximum value across states with non-zero probability
min_voi(): Minimum value across states with non-zero probability
stdev_voi(): Probability-weighted standard deviation (negated so maximizing = less spread)
quantile_voi(): Quantile of a discrete probability distribution

CRRA and CARA utility and inverse-utility transformations.

Tools for evaluating elicitation performance.

run_accuracy_vs_questions(): Simulate elicitation accuracy across question counts
run_asymptotic_performance_suite(): Run the asymptotic calibration suite for the elicitation algorithm

Published decision problems included for worked examples and validation.

canessa2015: Canessa et al. (2015) chytrid frog conservation problem
runge2011: Runge et al. (2011) fire ant eradication problem
runge2011_crane: Whooping Crane EVPXI problem from Runge et al. (2011)
turtle: Turtle reintroduction problem
williams2015: Williams et al. (2015) predator-prey management problem
bennett2018_parcel: Species Protection (Single Parcel & Two-Parcel Budget) from Bennett et al. (2018)
bennett2018_multispecies: Multi-Species Triage from Bennett et al. (2018)
mantyniemi2009_fishing: Fishing Location Choice and North Sea Herring from Mäntyniemi et al. (2009)
davis2019_s1: Davis et al. (2019) System 1 pre-computed performance matrices