Title:	Visualizing and Quantifying Decision Uncertainty
Version:	1.0.0
Description:	A suite of tools to help modelers and decision-makers effectively interpret and communicate decision risk when evaluating multiple policy options. It uses model outputs from uncertainty analysis for baseline scenarios and policy alternatives to generate visual representations of uncertainty and quantitative measures for assessing associated risks. For more details see https://dut.ihe.ca/.
License:	GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
Encoding:	UTF-8
RoxygenNote:	7.3.1
Suggests:	DT, knitr, rmarkdown, testthat (≥ 3.0.0)
VignetteBuilder:	knitr
Depends:	R (≥ 3.5.0)
LazyData:	true
LazyDataCompression:	xz
Imports:	ggdist, ggplot2, ggtext, rlang, scales, stats
Config/testthat/edition:	3
NeedsCompilation:	no
Packaged:	2025-09-09 15:47:02 UTC; mwiggins
Author:	Megan Wiggins [aut, cre], Marie Betsy Varughese [aut], Ellen Rafferty [aut], Sasha van Katwyk [aut], Christopher McCabe [aut], Jeff Round [aut], Erin Kirwin [aut], Institute of Health Economics [cph, aut], Canadian Network for Modelling Infectious Diseases [fnd]
Maintainer:	Megan Wiggins <mwiggins@ihe.ca>
Repository:	CRAN
Date/Publication:	2025-09-14 16:10:18 UTC

Calculate risk measures at peak (or minimum)

Description

For each policy alternative, this function calculates the risk measure at the peak values (or lowest values if the threshold is a minimum).

Usage

calculate_max_min_risk(max_min_values_list, D, Dt_max = TRUE)

Arguments

Value

A list of risk measure values.

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
D <- 750

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

calculate_max_min_risk(
  peak_values_list,
  D = D,
  Dt_max = TRUE
)

Calculate risk measures

Description

This function calculates the expected risk measure for each policy alternative using the outputs from multiple model runs with different input parameter sets (e.g., probabilistic sensitivity, uncertainty, or Bayesian inference analysis).

Usage

calculate_risk(data, tmin, tmax, Dt, Dt_max = TRUE, W = FALSE, weight = NULL)

Arguments

Value

A list of risk scores, one for each policy alternative.

data format

Each data.frame in data contains the results from multiple model runs using different parameter sets (e.g., from probabilistic sensitivity, uncertainty, or Bayesian inference analysis). The first column contains the model time and subsequent columns contain the predicted output for each simulation run at the respective time. The model time in the first column must contain numeric values indicating a simulation time (ex. 1, 2, 3,...) or dates (ex. "2021-01-01", "2021-01-02") which must be in R Date format (i.e., class="Date"). To ensure a consistent basis for comparison, the model time in the first column should be the same for each policy alternative (data.frame).

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
Dt <- rep(750, nrow(psa_data$Baseline))

calculate_risk(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt = Dt,
  Dt_max = TRUE
)

Calculate threshold exceedance probabilities

Description

For each policy alternative, this function calculates the probability that the peak (or minimum) value exceeds (or is below) a specified threshold(s) using a Riemann sum approach.

Usage

calculate_threshold_probs(max_min_values_list, Dp, Dt_max = TRUE)

Arguments

Value

A list of vectors of threshold values and corresponding probabilities.

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
D <- 750
Dp <- c(750, 1000, 2000)

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

calculate_threshold_probs(
  peak_values_list,
  Dp = Dp,
  Dt_max = TRUE
)

Calculate time of threshold exceedance

Description

For each policy alternative, this function calculates: (i) the percent of simulations in which the threshold is exceeded (or not met if the threshold is a minimum), (ii) the mean simulation time of the first exceedance and 95th percentile range, (iii) the mean duration of the first exceedance and 95th percentile range, and (iv) if the first column of model output data.frame passed to the function is a Date, the mean date of the first and last exceedance.

Usage

calculate_time(data, tmin, tmax, Dt, Dt_max = TRUE)

Arguments

Value

A list of data.frames, one for each policy alternative.

data format

Each data.frame in data contains the results from multiple model runs using different parameter sets (e.g., from probabilistic sensitivity, uncertainty, or Bayesian inference analysis). The first column contains the model time and subsequent columns contain the predicted output for each simulation run at the respective time. The model time in the first column must contain numeric values indicating a simulation time (ex. 1, 2, 3,...) or dates (ex. "2021-01-01", "2021-01-02") which must be in R Date format (i.e., class="Date"). To ensure a consistent basis for comparison, the model time in the first column should be the same for each policy alternative (data.frame).

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
Dt <- rep(750, nrow(psa_data$Baseline))

calculate_time(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt = Dt,
  Dt_max = TRUE
)

Finds peak (or lowest) model output values

Description

For each policy alternative, this function finds the peak (or lowest if the threshold is a minimum) model output values for each simulation run.

Usage

get_max_min_values(data, tmin, tmax, Dt_max = TRUE)

Arguments

Value

A list of data.frame(s) containing the peak (or lowest) value and corresponding simulation time for each policy alternative.

data format

Each data.frame in data contains the results from multiple model runs using different parameter sets (e.g., from probabilistic sensitivity, uncertainty, or Bayesian inference analysis). The first column contains the model time and subsequent columns contain the predicted output for each simulation run at the respective time. The model time in the first column must contain numeric values indicating a simulation time (ex. 1, 2, 3,...) or dates (ex. "2021-01-01", "2021-01-02") which must be in R Date format (i.e., class="Date"). To ensure a consistent basis for comparison, the model time in the first column should be the same for each policy alternative (data.frame).

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
Dt <- rep(750, nrow(psa_data$Baseline))

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

Find output values at time points relative to the peak (or minimum)

Description

For each policy alternative, this function finds the the model output values for each simulation run at the specified time points relative to the peak (or minimum) value.

Usage

get_relative_values(data, max_min_values_list, t_s, t_ss, Dt_max = TRUE)

Arguments

Value

A list of nested lists with the following elements: (i) a data.frame of values recorded at specific time steps relative to the peak time (e.g., peak-10, peak, peak+10) for each model run, and (ii) a vector of the names of each time step ex. minus_30,minus_20,...,plus_20, plus_30.

data format

Each data.frame in data contains the results from multiple model runs using different parameter sets (e.g., from probabilistic sensitivity, uncertainty, or Bayesian inference analysis). The first column contains the model time and subsequent columns contain the predicted output for each simulation run at the respective time. The model time in the first column must contain numeric values indicating a simulation time (ex. 1, 2, 3,...) or dates (ex. "2021-01-01", "2021-01-02") which must be in R Date format (i.e., class="Date"). To ensure a consistent basis for comparison, the model time in the first column should be the same for each policy alternative (data.frame).

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
D <- 750
t_s <- 20
t_ss <- 10

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

peak_temporal_list <- get_relative_values(
  psa_data,
  peak_values_list,
  t_s = t_s,
  t_ss = t_ss
)

Generate density plots

Description

For each policy alternative, this function generates probability density plots of the highest (or lowest if the threshold is a minimum) projected outcome across simulation runs. The decision threshold is shown directly on the plot as a vertical line. The area under the probability density curve where the threshold value is exceeded is shaded to visually display the downside risk of the policy alternative.

Usage

plot_density(max_min_values_list, D, Dt_max = TRUE, risk_measures)

Arguments

Value

A list of ggplots, one for each policy alternative.

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
Dt <- rep(750, nrow(psa_data$Baseline))
D <- 750

risk_measures <- calculate_risk(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt = Dt,
  Dt_max = TRUE
)

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

density_plots <- plot_density(
  peak_values_list,
  D = D,
  Dt_max = TRUE,
  risk_measures
)

Generate time-outcome fan plots

Description

For each policy alternative, this function generates time-outcome fan plots summarizing the trajectory of the outcome over time using the mean value for a given policy alternative. Uncertainty is characterized by shading the 50% and 95% credible intervals (calculated as 25th and 75th percentiles and 2.5th and 97.5th percentiles, respectively). The decision threshold is shown directly on the plot to provide a clear reference point for interpreting the outcome values.

Usage

plot_fan(data, tmin, tmax, Dt, Dt_max = TRUE)

Arguments

Value

A list of ggplots, one for each policy alternative.

data format

Each data.frame in data contains the results from multiple model runs using different parameter sets (e.g., from probabilistic sensitivity, uncertainty, or Bayesian inference analysis). The first column contains the model time and subsequent columns contain the predicted output for each simulation run at the respective time. The model time in the first column must contain numeric values indicating a simulation time (ex. 1, 2, 3,...) or dates (ex. "2021-01-01", "2021-01-02") which must be in R Date format (i.e., class="Date"). To ensure a consistent basis for comparison, the model time in the first column should be the same for each policy alternative (data.frame).

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
Dt <- rep(750, nrow(psa_data$Baseline))

fan_plots <- plot_fan(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt = Dt,
  Dt_max = TRUE
)

Generate raincloud plot

Description

The probability densities of the highest (or lowest if the threshold is a minimum) projected outcome across simulation runs are plotted for each policy alternative alongside corresponding box plots, which indicate the mean and 50th percentile range. These plots are presented collectively on a single graph to facilitate visual comparison of the policy alternatives. The decision threshold is shown directly on the plot as a vertical line to provide a clear reference point for interpreting the outputs.

Usage

plot_raincloud(max_min_values_list, D)

Arguments

Value

A raincloud plot for peak (or minimum) values

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
D <- 750

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

plot_raincloud(
  peak_values_list,
  D = D
)

Generate temporal probability density plots

Description

The probability density of the peak (or lowest if the threshold is a minimum) projected outcome across simulation runs is plotted in the center of the graph for a given policy alternative. Above and below, the probability density of the outcome at specified time points relative to the time of the peak (or lowest) project outcome is plotted to visually illustrate how uncertainty, and therefore risk, changes over time. The decision threshold is shown directly on the plot as a vertical line to provide a clear reference point for interpreting the outputs.

Usage

plot_temporal(relative_values, D)

Arguments

Value

A list of ggplots, one for each policy alternative.

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
D <- 750
t_s <- 30
t_ss <- 10

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

peak_temporal_list <- get_relative_values(
  psa_data,
  peak_values_list,
  t_s = t_s,
  t_ss = t_ss
)

peak_temporal_plots <- plot_temporal(
  peak_temporal_list,
  D
)

Sample PSA data

Description

Sample simulation outputs for a baseline scenario and three interventions, in a format compatible with package functions

Usage

psa_data

Format

psa_data

A list of data frames, one for each scenario corresponding to the list item name (Baseline, Intervention_1). Each data frame has 100 rows and 814 columns:

date

(Baseline only) the simulation date in calendar time

time

(Intervention scenarios only) simulation time as integer starting with 0

X1, ..., X1727

individual simulation results

Calculates summary statistics at specified time points relative to the peak (or minimum)

Description

For each policy alternative, this function calculates summary statistics (n, mean, median, and IQR) for the model output values at the specified time points relative to the peak (or minimum) value.

Usage

sum_stats_temporal(relative_values)

Arguments

Value

A list of data.frames of summary statistics.

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
D <- 750
t_s <- 20
t_ss <- 10

peak_values_list <- get_max_min_values(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt_max = TRUE
)

peak_temporal_list <- get_relative_values(
  psa_data,
  peak_values_list,
  t_s = t_s,
  t_ss = t_ss
)

stats_peak_temporal <- sum_stats_temporal(peak_temporal_list)

Output risk measures to table

Description

This function tabulates the expected risk values for each policy alternative and the percent change in risk relative to the baseline scenario.

Usage

tabulate_risk(exp_risk_list, n_s)

Arguments

Value

A character matrix of risk values and policy risk impact (%) for each policy alternative.

Examples

tmin <- "2021-01-01"
tmax <- "2021-04-10"
Dt <- rep(750, nrow(psa_data$Baseline))

risk_measures <- calculate_risk(
  psa_data,
  tmin = tmin,
  tmax = tmax,
  Dt = Dt,
  Dt_max = TRUE
)

tabulate_risk(
  risk_measures,
  n_s = length(psa_data)
)