| Title: | Alert Time Quality - Evaluating Timely Epidemic Metrics |
|---|---|
| Description: | Provides tools for evaluating timely epidemic detection models within school absenteeism-based surveillance systems. Introduces the concept of alert time quality as an evaluation metric. Includes functions to simulate populations, epidemics, and alert metrics associated with epidemic spread using population census data. The methods are based on research published in Vanderkruk et al. (2023) <doi:10.1186/s12889-023-15747-z> and Ward et al. (2019) <doi:10.1186/s12889-019-7521-7>. |
| Authors: | Vinay Joshy [aut, cre, cph], Zeny Feng [aut, cph, ths], Lorna Deeth [aut, cph, ths], Justin Slater [aut, cph, ths], Kayla Vanderkruk [aut, com, ctb] |
| Maintainer: | Vinay Joshy <[email protected]> |
| License: | GPL (>= 3) |
| Version: | 0.2.3 |
| Built: | 2024-09-17 05:30:07 UTC |
| Source: | https://github.com/vjoshy/atq_surveillance_package |
This function creates an object of class "alarm_metrics" from a list of metric matrices. It's used to organize and structure the results from various epidemic alarm metrics calculations.
alarm_metrics(metrics_list)alarm_metrics(metrics_list)
metrics_list |
A list containing matrices for different alarm metrics (FAR, ADD, AATQ, FATQ, WAATQ, WFATQ). |
An object of class "alarm_metrics" which is a list with the input metrics.
# Generate sample alarm metrics data set.seed(123) generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix() ) # Create an alarm_metrics object alarm_metrics_obj <- alarm_metrics(sample_metrics) # Check the class of the resulting object class(alarm_metrics_obj) # Access a specific metric head(alarm_metrics_obj$AATQ) # Use with other functions (assuming they're defined in your package) # plot(alarm_metrics_obj, metric = "FAR")# Generate sample alarm metrics data set.seed(123) generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix() ) # Create an alarm_metrics object alarm_metrics_obj <- alarm_metrics(sample_metrics) # Check the class of the resulting object class(alarm_metrics_obj) # Access a specific metric head(alarm_metrics_obj$AATQ) # Use with other functions (assuming they're defined in your package) # plot(alarm_metrics_obj, metric = "FAR")
This function creates an object of class "alarm_plot_data" from epidemic data and best models. It ensures that the input data is in the correct format for plotting.
alarm_plot_data(epidemic_data, best_models)alarm_plot_data(epidemic_data, best_models)
epidemic_data |
A data frame or list of data frames containing epidemic data. |
best_models |
A list of data frames containing the best models for each metric. |
An object of class "alarm_plot_data" containing the epidemic data and best models.
# Generate sample epidemic data set.seed(123) epidemic_data <- data.frame( Date = rep(1:300, 3), ScYr = rep(1:3, each = 300), pct_absent = runif(900, 0, 0.1), lab_conf = rpois(900, lambda = 5), ref_date = rep(c(rep(0, 299), 1), 3), Case = rbinom(900, 1, 0.1), window = rep(c(rep(0, 280), rep(1, 20)), 3) ) # Generate sample best models data generate_best_model <- function() { data.frame( Date = sample(1:300, 50), ScYr = sample(1:3, 50, replace = TRUE), Alarm = sample(0:1, 50, replace = TRUE), lag = sample(1:15, 50, replace = TRUE), thres = runif(50, 0, 1) ) } best_models <- list( AATQ = generate_best_model(), FATQ = generate_best_model(), FAR = generate_best_model(), ADD = generate_best_model(), WFATQ = generate_best_model(), WAATQ = generate_best_model() ) # Create alarm_plot_data object plot_data <- alarm_plot_data(epidemic_data, best_models) # Check the structure of the resulting object str(plot_data)# Generate sample epidemic data set.seed(123) epidemic_data <- data.frame( Date = rep(1:300, 3), ScYr = rep(1:3, each = 300), pct_absent = runif(900, 0, 0.1), lab_conf = rpois(900, lambda = 5), ref_date = rep(c(rep(0, 299), 1), 3), Case = rbinom(900, 1, 0.1), window = rep(c(rep(0, 280), rep(1, 20)), 3) ) # Generate sample best models data generate_best_model <- function() { data.frame( Date = sample(1:300, 50), ScYr = sample(1:3, 50, replace = TRUE), Alarm = sample(0:1, 50, replace = TRUE), lag = sample(1:15, 50, replace = TRUE), thres = runif(50, 0, 1) ) } best_models <- list( AATQ = generate_best_model(), FATQ = generate_best_model(), FAR = generate_best_model(), ADD = generate_best_model(), WFATQ = generate_best_model(), WAATQ = generate_best_model() ) # Create alarm_plot_data object plot_data <- alarm_plot_data(epidemic_data, best_models) # Check the structure of the resulting object str(plot_data)
This function sets the class of the given model data to "best_model".
best_model(model_data)best_model(model_data)
model_data |
A data frame containing the model data. |
An object of class "best_model" containing the model data.
# Generate sample model data sample_model_data <- data.frame( ScYr = rep(1:3, each = 100), Date = rep(1:100, 3), Alarm = sample(c(0, 1), 300, replace = TRUE, prob = c(0.9, 0.1)), lag = sample(1:15, 300, replace = TRUE), thres = runif(300, 0.1, 0.6) ) # Create best model best_model_data <- best_model(sample_model_data) # Print the class of the best model class(best_model_data)# Generate sample model data sample_model_data <- data.frame( ScYr = rep(1:3, each = 100), Date = rep(1:100, 3), Alarm = sample(c(0, 1), 300, replace = TRUE, prob = c(0.9, 0.1)), lag = sample(1:15, 300, replace = TRUE), thres = runif(300, 0.1, 0.6) ) # Create best model best_model_data <- best_model(sample_model_data) # Print the class of the best model class(best_model_data)
Function to simulate specified catchments of square area (a x a). The number of schools in each catchment area is simulated via a specified distribution function, with gamma distribution as the default.
catchment_sim(n, area, dist_func = stats::rgamma, ...)catchment_sim(n, area, dist_func = stats::rgamma, ...)
n |
number of catchments to be simulated |
area |
square dimension for catchment (if area = 20, then each catchment dimensions will be 20 x 20) |
dist_func |
distribution function to simulate number of schools, default is stats::rgamma |
... |
additional arguments passed to the distribution function |
A data frame with n rows and the following columns:
catchID |
Unique identifier for each catchment area |
num.schools |
Number of schools in the catchment area |
xStart |
Starting x-coordinate of the catchment area |
xEnd |
Ending x-coordinate of the catchment area |
yStart |
Starting y-coordinate of the catchment area |
yEnd |
Ending y-coordinate of the catchment area |
# Using default gamma distribution catch_df1 <- catchment_sim(16, 20, shape = 4.1, rate = 2.7) # Using normal distribution catch_df2 <- catchment_sim(16, 20, dist_func = stats::rnorm, mean = 5, sd = 1) # Using Poisson distribution catch_df3 <- catchment_sim(16, 20, dist_func = stats::rpois, lambda = 3)# Using default gamma distribution catch_df1 <- catchment_sim(16, 20, shape = 4.1, rate = 2.7) # Using normal distribution catch_df2 <- catchment_sim(16, 20, dist_func = stats::rnorm, mean = 5, sd = 1) # Using Poisson distribution catch_df3 <- catchment_sim(16, 20, dist_func = stats::rpois, lambda = 3)
Compiles and processes epidemic data, simulating school absenteeism using epidemic and individual data.
compile_epi(epidemic, individual_data, lags = 16, inf_period = 4, T = 300)compile_epi(epidemic, individual_data, lags = 16, inf_period = 4, T = 300)
epidemic |
A list of simulated epidemic data (output from simepi() or similar function). |
individual_data |
A data frame of simulated individuals data (output from simulate_households() or similar function). |
lags |
Maximum number of lags for absenteeism, default = 16 (note that lag of zero is included). |
inf_period |
Infection period of epidemic, default = 4. |
T |
Time period, default = 300 days. |
A data frame containing compiled epidemic data, including:
Date |
Day of the epidemic |
ScYr |
School year |
pct_absent |
Percentage of students absent |
absent |
Number of students absent |
absent_sick |
Number of students absent due to illness |
new_inf |
Number of new infections |
reported_cases |
Number of reported cases |
Case |
Indicator for lab confirmed flu case that day |
sinterm, costerm
|
Seasonal terms |
window |
Indicator for "True Alarm" window |
ref_date |
Indicator for reference date |
lag0, lag1, ..., lag15
|
Lagged absenteeism values |
# Assuming you have previously simulated epidemic and individual data: epidemic <- ssir(1000, alpha = 0.4) individual_data <- data.frame(elem_child_ind = sample(0:1, 1000, replace = TRUE), schoolID = sample(1:10, 1000, replace = TRUE)) compiled_data <- compile_epi(epidemic, individual_data)# Assuming you have previously simulated epidemic and individual data: epidemic <- ssir(1000, alpha = 0.4) individual_data <- data.frame(elem_child_ind = sample(0:1, 1000, replace = TRUE), schoolID = sample(1:10, 1000, replace = TRUE)) compiled_data <- compile_epi(epidemic, individual_data)
This function creates a summary of alarm metrics, including statistics for each metric, best model parameters, reference dates, and best prediction dates for each epidemic year.
create_alarm_metrics_summary(metrics, best_models, epidemic_data)create_alarm_metrics_summary(metrics, best_models, epidemic_data)
metrics |
An object of class "alarm_metrics" containing matrices for different metrics. |
best_models |
A list of data frames, each containing the best model for a specific metric. |
epidemic_data |
A data frame containing the epidemic data, including ScYr, Date, and ref_date columns. |
An object of class "alarm_metrics_summary" containing summary statistics, best values, reference dates, and best prediction dates for each metric and epidemic year.
# Generate sample data set.seed(123) # Generate sample metrics generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix() ) metrics <- structure(sample_metrics, class = c("alarm_metrics", "list")) # Generate sample best models generate_best_model <- function() { data.frame( ScYr = rep(1:3, each = 100), Date = rep(1:100, 3), Alarm = sample(c(0, 1), 300, replace = TRUE, prob = c(0.9, 0.1)), lag = sample(1:15, 300, replace = TRUE), thres = runif(300, 0.1, 0.6) ) } best_models <- list( best.FAR = generate_best_model(), best.ADD = generate_best_model(), best.AATQ = generate_best_model(), best.FATQ = generate_best_model(), best.WAATQ = generate_best_model(), best.WFATQ = generate_best_model() ) # Generate sample epidemic data epidemic_data <- data.frame( ScYr = rep(1:3, each = 365), Date = rep(1:365, 3), ref_date = c(rep(0, 364), 1) ) # Create alarm metrics summary summary <- create_alarm_metrics_summary(metrics, best_models, epidemic_data) # Print summary print(summary)# Generate sample data set.seed(123) # Generate sample metrics generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix() ) metrics <- structure(sample_metrics, class = c("alarm_metrics", "list")) # Generate sample best models generate_best_model <- function() { data.frame( ScYr = rep(1:3, each = 100), Date = rep(1:100, 3), Alarm = sample(c(0, 1), 300, replace = TRUE, prob = c(0.9, 0.1)), lag = sample(1:15, 300, replace = TRUE), thres = runif(300, 0.1, 0.6) ) } best_models <- list( best.FAR = generate_best_model(), best.ADD = generate_best_model(), best.AATQ = generate_best_model(), best.FATQ = generate_best_model(), best.WAATQ = generate_best_model(), best.WFATQ = generate_best_model() ) # Generate sample epidemic data epidemic_data <- data.frame( ScYr = rep(1:3, each = 365), Date = rep(1:365, 3), ref_date = c(rep(0, 364), 1) ) # Create alarm metrics summary summary <- create_alarm_metrics_summary(metrics, best_models, epidemic_data) # Print summary print(summary)
Function to simulate elementary school size and assigns it to catchments. The school population is simulated using a specified distribution function, with gamma distribution as the default.
elementary_pop(df, dist_func = stats::rgamma, ...)elementary_pop(df, dist_func = stats::rgamma, ...)
df |
output data frame from catchment_sim function |
dist_func |
distribution function to simulate school population, default is stats::rgamma |
... |
additional arguments passed to the distribution function |
A data frame with the following columns:
catchID |
Identifier for the catchment area |
schoolID |
Unique identifier for each school |
schoolPop |
Simulated population of the school |
xStart |
Starting x-coordinate of the catchment area |
xEnd |
Ending x-coordinate of the catchment area |
yStart |
Starting y-coordinate of the catchment area |
yEnd |
Ending y-coordinate of the catchment area |
# Simulate catchment areas catch_df <- catchment_sim(16, 20, shape = 3.5, rate = 2.8) # Simulate elementary schools using default gamma distribution elementary_df1 <- elementary_pop(catch_df, shape = 5.1, rate = 0.015) # Simulate elementary schools using normal distribution elementary_df2 <- elementary_pop(catch_df, dist_func = stats::rnorm, mean = 300, sd = 50) # Simulate elementary schools using Poisson distribution elementary_df3 <- elementary_pop(catch_df, dist_func = stats::rpois, lambda = 250)# Simulate catchment areas catch_df <- catchment_sim(16, 20, shape = 3.5, rate = 2.8) # Simulate elementary schools using default gamma distribution elementary_df1 <- elementary_pop(catch_df, shape = 5.1, rate = 0.015) # Simulate elementary schools using normal distribution elementary_df2 <- elementary_pop(catch_df, dist_func = stats::rnorm, mean = 300, sd = 50) # Simulate elementary schools using Poisson distribution elementary_df3 <- elementary_pop(catch_df, dist_func = stats::rpois, lambda = 250)
This function calculates various performance metrics for epidemic alarm systems across different lags and thresholds. It evaluates False Alarm Rate (FAR), Added Days Delayed (ADD), Average Alarm Time Quality (AATQ), First Alarm Time Quality (FATQ), and their weighted versions (WAATQ, WFATQ).
eval_metrics(data, maxlag = 15, thres = seq(0.1, 0.6, by = 0.05), topt = 14)eval_metrics(data, maxlag = 15, thres = seq(0.1, 0.6, by = 0.05), topt = 14)
data |
A data frame containing the epidemic data with lagged variables, typically output from the compile_epi function. |
maxlag |
The maximum number of lags to consider (default: 15). |
thres |
A vector of threshold values to evaluate (default: seq(0.1, 0.6, by = 0.05)). |
topt |
Optimal alarm day (default = 14). |
A list containing three elements:
metrics |
An object of class "alarm_metrics" with the following components:
|
plot_data |
An object of class "alarm_plot_data" for generating plots |
summary |
An object of class "alarm_metrics_summary" containing summary statistics |
# Generate simulated epidemic data n_rows <- 7421 n_houses <- 1000 epidemic_new <- ssir(n_rows, T = 300, alpha = 0.298, inf_init = 32, rep = 3) individual_data <- data.frame( houseID = rep(1:n_houses, each = ceiling(n_rows / n_houses))[1:n_rows], catchID = sample(1:10, n_rows, replace = TRUE), schoolID = sample(1:10, n_rows, replace = TRUE), num_people = round(rnorm(n_rows, mean = 4, sd = 1)), num_elem_child = round(rnorm(n_rows, mean = 1, sd = 1)), xStart = 0, xEnd = 5, yStart = 0, yEnd = 5, loc.x = rnorm(n_rows, mean = 2.5, sd = 1), loc.y = rnorm(n_rows, mean = 2.5, sd = 1), individualID = 1:n_rows, elem_child_ind = sample(0:1, n_rows, replace = TRUE) ) compiled_data <- compile_epi(epidemic_new, individual_data) # Evaluate alarm metrics alarm_metrics <- eval_metrics(compiled_data, thres = seq(0.1, 0.3, by = 0.05)) # Access the results summary(alarm_metrics$summary)# Generate simulated epidemic data n_rows <- 7421 n_houses <- 1000 epidemic_new <- ssir(n_rows, T = 300, alpha = 0.298, inf_init = 32, rep = 3) individual_data <- data.frame( houseID = rep(1:n_houses, each = ceiling(n_rows / n_houses))[1:n_rows], catchID = sample(1:10, n_rows, replace = TRUE), schoolID = sample(1:10, n_rows, replace = TRUE), num_people = round(rnorm(n_rows, mean = 4, sd = 1)), num_elem_child = round(rnorm(n_rows, mean = 1, sd = 1)), xStart = 0, xEnd = 5, yStart = 0, yEnd = 5, loc.x = rnorm(n_rows, mean = 2.5, sd = 1), loc.y = rnorm(n_rows, mean = 2.5, sd = 1), individualID = 1:n_rows, elem_child_ind = sample(0:1, n_rows, replace = TRUE) ) compiled_data <- compile_epi(epidemic_new, individual_data) # Evaluate alarm metrics alarm_metrics <- eval_metrics(compiled_data, thres = seq(0.1, 0.3, by = 0.05)) # Access the results summary(alarm_metrics$summary)
This function creates two plots for a single epidemic simulation: one for new infections and one for reported cases.
plot_single_epidemic(epidemic, epidemic_num)plot_single_epidemic(epidemic, epidemic_num)
epidemic |
A list containing the results of a single epidemic simulation, typically an element from an "ssir_epidemic_multi" object. |
epidemic_num |
An integer indicating the epidemic number, used in the plot titles. |
A grid arrangement of two ggplot objects: new infections and reported cases.
# Assuming you have run a simulation: result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) plot_single_epidemic(result, 1) # For multiple simulations: multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 5) plot_single_epidemic(multi_result[[1]], 1)# Assuming you have run a simulation: result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) plot_single_epidemic(result, 1) # For multiple simulations: multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 5) plot_single_epidemic(multi_result[[1]], 1)
This function creates a heatmap visualization of the specified alarm metric across different lags and thresholds.
## S3 method for class 'alarm_metrics' plot(x, metric = "AATQ", col = heat.colors(12), ...)## S3 method for class 'alarm_metrics' plot(x, metric = "AATQ", col = heat.colors(12), ...)
x |
An object of class "alarm_metrics" containing matrices of metric values. |
metric |
A character string specifying which metric to plot. Default is "AATQ". Options include "FAR", "ADD", "AATQ", "FATQ", "WAATQ", "WFATQ". |
col |
Set heat map color profile, default is heat.colors(15) |
... |
Additional arguments passed to the image function. |
A heatmap plot of the specified metric.
# Generate sample alarm metrics data set.seed(123) generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix(), lags = 1:15, thresholds = seq(0.1, 0.6, by = 0.05) ) # Create an alarm_metrics object alarm_metrics_obj <- structure(sample_metrics, class = c("alarm_metrics", "list")) # Plot the heatmap for AATQ (default) plot(alarm_metrics_obj) # Plot the heatmap for FAR plot(alarm_metrics_obj, metric = "FAR") # Customize the plot plot(alarm_metrics_obj, metric = "FATQ", col = heat.colors(12))# Generate sample alarm metrics data set.seed(123) generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix(), lags = 1:15, thresholds = seq(0.1, 0.6, by = 0.05) ) # Create an alarm_metrics object alarm_metrics_obj <- structure(sample_metrics, class = c("alarm_metrics", "list")) # Plot the heatmap for AATQ (default) plot(alarm_metrics_obj) # Plot the heatmap for FAR plot(alarm_metrics_obj, metric = "FAR") # Customize the plot plot(alarm_metrics_obj, metric = "FATQ", col = heat.colors(12))
This function creates a series of plots, one for each epidemic year, showing the absenteeism percentage, laboratory confirmed cases, reference dates, and alarm points for different models.
## S3 method for class 'alarm_plot_data' plot(x, ...)## S3 method for class 'alarm_plot_data' plot(x, ...)
x |
An object of class "alarm_plot_data" containing epidemic data and best models. |
... |
Additional arguments passed to the plotting function (currently unused). |
A list of ggplot objects, one for each epidemic year.
# Generate sample data # Generate simulated epidemic data n_rows <- 7421 n_houses <- 1000 epidemic_new <- ssir(n_rows, T = 300, alpha =0.298, inf_init = 32, rep = 3) individual_data <- data.frame( houseID = rep(1:n_houses, each = ceiling(n_rows / n_houses))[1:n_rows], catchID = sample(1:10, n_rows, replace = TRUE), schoolID = sample(1:10, n_rows, replace = TRUE), num_people = round(rnorm(n_rows, mean = 4, sd = 1)), # Normal distribution for num_people num_elem_child = round(rnorm(n_rows, mean = 1, sd = 1)), # Normal distribution for num_elem_child xStart = 0, xEnd = 5, yStart = 0, yEnd = 5, loc.x = rnorm(n_rows, mean = 2.5, sd = 1), # Normal distribution for loc.x loc.y = rnorm(n_rows, mean = 2.5, sd = 1), # Normal distribution for loc.y individualID = 1:n_rows, elem_child_ind = sample(0:1, n_rows, replace = TRUE) ) compiled_data <- compile_epi(epidemic_new, individual_data) # Evaluate alarm metrics results <- eval_metrics(compiled_data, thres = seq(0.1,0.6,by = 0.05)) # Generate plots plots <- plot(results$plot_data) # Display the first plot if(interactive()) print(plots[[1]])# Generate sample data # Generate simulated epidemic data n_rows <- 7421 n_houses <- 1000 epidemic_new <- ssir(n_rows, T = 300, alpha =0.298, inf_init = 32, rep = 3) individual_data <- data.frame( houseID = rep(1:n_houses, each = ceiling(n_rows / n_houses))[1:n_rows], catchID = sample(1:10, n_rows, replace = TRUE), schoolID = sample(1:10, n_rows, replace = TRUE), num_people = round(rnorm(n_rows, mean = 4, sd = 1)), # Normal distribution for num_people num_elem_child = round(rnorm(n_rows, mean = 1, sd = 1)), # Normal distribution for num_elem_child xStart = 0, xEnd = 5, yStart = 0, yEnd = 5, loc.x = rnorm(n_rows, mean = 2.5, sd = 1), # Normal distribution for loc.x loc.y = rnorm(n_rows, mean = 2.5, sd = 1), # Normal distribution for loc.y individualID = 1:n_rows, elem_child_ind = sample(0:1, n_rows, replace = TRUE) ) compiled_data <- compile_epi(epidemic_new, individual_data) # Evaluate alarm metrics results <- eval_metrics(compiled_data, thres = seq(0.1,0.6,by = 0.05)) # Generate plots plots <- plot(results$plot_data) # Display the first plot if(interactive()) print(plots[[1]])
This function creates plots for a single SSIR (Stochastic Susceptible-Infectious-Removed) epidemic simulation. It generates two plots: one for new infections and one for reported cases.
## S3 method for class 'ssir_epidemic' plot(x, ...)## S3 method for class 'ssir_epidemic' plot(x, ...)
x |
An object of class "ssir_epidemic", typically the result of calling ssir() with rep = NULL or 1. |
... |
Additional arguments passed to the underlying plotting function (currently unused). |
A grid arrangement of two ggplot objects: new infections and reported cases.
plot_single_epidemic for the underlying plotting function
# Run a single simulation result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) # Plot the results plot(result)# Run a single simulation result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) # Plot the results plot(result)
This function creates plots for multiple SSIR (Stochastic Susceptible-Infectious-Removed) epidemic simulations. It generates two plots (new infections and reported cases) for each simulation in the multi-simulation object.
## S3 method for class 'ssir_epidemic_multi' plot(x, ...)## S3 method for class 'ssir_epidemic_multi' plot(x, ...)
x |
An object of class "ssir_epidemic_multi", typically the result of calling ssir() with rep > 1. |
... |
Additional arguments passed to the underlying plotting function (currently unused). |
A list of grid arrangements, each containing two ggplot objects (new infections and reported cases) for each epidemic simulation.
plot_single_epidemic for the underlying plotting function
# Run multiple simulations multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 5) # Plot the results plots <- plot(multi_result) # Display the first simulation's plots plots[[1]]# Run multiple simulations multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 5) # Plot the results plots <- plot(multi_result) # Display the first simulation's plots plots[[1]]
Creates two simulated populations, households and individuals
simulate_households(children_df, noChildren_df)simulate_households(children_df, noChildren_df)
children_df |
data frame output from house_children() function |
noChildren_df |
data frame output from house_noChildren() function |
list of two data frames; simulated individuals data and simulated households data
# Simulate catchment areas catch_df <- catchment_sim(4, 5, shape = 2.5, rate = 1.8) # Simulate elementary schools using default gamma distribution elementary_df <- elementary_pop(catch_df, shape = 4.1, rate = 0.019) # Simulate households with children house_children <- subpop_children(elementary_df, n = 2, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.2) # Simulate households without children using pre-specified proportions house_nochildren <- subpop_noChildren(house_children, elementary_df, prop_house_size = c(0.2, 0.3, 0.25, 0.15, 0.1), prop_house_Children = 0.3) # simulate households and individuals data simulation <- simulate_households(house_children, house_nochildren)# Simulate catchment areas catch_df <- catchment_sim(4, 5, shape = 2.5, rate = 1.8) # Simulate elementary schools using default gamma distribution elementary_df <- elementary_pop(catch_df, shape = 4.1, rate = 0.019) # Simulate households with children house_children <- subpop_children(elementary_df, n = 2, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.2) # Simulate households without children using pre-specified proportions house_nochildren <- subpop_noChildren(house_children, elementary_df, prop_house_size = c(0.2, 0.3, 0.25, 0.15, 0.1), prop_house_Children = 0.3) # simulate households and individuals data simulation <- simulate_households(house_children, house_nochildren)
This function runs a stochastic Susceptible-Infectious-Removed (SIR) epidemic simulation. It can run a single simulation or multiple replications. Also simulates laboratory confirmed cases
ssir( N, T = 300, alpha, avg_start = 45, min_start = 20, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = NULL )ssir( N, T = 300, alpha, avg_start = 45, min_start = 20, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = NULL )
N |
Numeric. The total population size. |
T |
Numeric. The duration of the simulation in time steps. Default is 300. |
alpha |
Numeric. The transmission rate. Must be a number between 0 and 1. |
avg_start |
Numeric. The average start day of the epidemic. Default is 45. |
min_start |
Numeric. The minimum start day of the epidemic. Default is 20. |
inf_period |
Numeric. The duration of the infectious period in time steps. Default is 4. |
inf_init |
Numeric. The initial number of infected individuals. Default is 32. |
report |
Numeric. The proportion of cases that are reported. Default is 0.02. |
lag |
Numeric. The average delay in reporting cases. Default is 7. |
rep |
Numeric or NULL. The number of simulation replications to run. If NULL or 1, a single simulation is run. |
If rep is NULL or 1, returns an object of class "ssir_epidemic" containing:
new_inf |
Vector of new infections at each time step |
reported_cases |
Vector of reported cases at each time step |
S |
Vector of susceptible individuals at each time step |
I |
Vector of infectious individuals at each time step |
R |
Vector of removed individuals at each time step |
parameters |
List of input parameters |
If rep > 1, returns an object of class "ssir_epidemic_multi" containing multiple simulation results.
# Run a single simulation result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) # Run multiple simulations multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 100)# Run a single simulation result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) # Run multiple simulations multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 100)
Simulation of households with children using specified random distributions. Number of observations is multiplied by a default value of five.
subpop_children( df, n = 5, prop_parent_couple = NULL, prop_children_couple = NULL, prop_children_lone = NULL, prop_elem_age = NULL, parent_dist = stats::runif, child_dist = stats::runif, age_dist = stats::runif, ... )subpop_children( df, n = 5, prop_parent_couple = NULL, prop_children_couple = NULL, prop_children_lone = NULL, prop_elem_age = NULL, parent_dist = stats::runif, child_dist = stats::runif, age_dist = stats::runif, ... )
df |
simulated output data frame from elementary_pop function |
n |
population multiplier, default value = 5 |
prop_parent_couple |
proportion of parents as a couple (optional) |
prop_children_couple |
vector of proportions for coupled parents with 1, 2, 3+ children (optional) |
prop_children_lone |
vector of proportions for single parents with 1, 2, 3+ children (optional) |
prop_elem_age |
proportion of children that are of elementary school age (optional) |
parent_dist |
distribution function for parent type, default is stats::runif |
child_dist |
distribution function for number of children, default is stats::runif |
age_dist |
distribution function for child age, default is stats::runif |
... |
additional arguments passed to the distribution functions |
This function can be used interactively or with pre-specified parameters. If proportions are not provided, the user will be prompted to enter them. Custom distribution functions can be specified for parent type, number of children, and child age. The total number of simulations (n * sum of school populations) is automatically passed as the first argument to each distribution function.
A data frame representing the simulated population of households with children, including:
schoolID |
Assigned school ID for the household |
houseID |
Unique identifier for each household |
num_parent |
Number of parents in the household (1 or 2) |
num_child |
Total number of children in the household |
num_elem_child |
Number of elementary school-aged children in the household |
catchID |
Assigned catchment ID for the household |
schoolPop |
Total population of elementary school assigned for the household |
xStart |
Starting X-coordindate for assigned catchment |
xEnd |
End X-coordindate for assigned catchment |
yStart |
Starting Y-coordindate for assigned catchment |
yEnd |
End Y-coordindate for assigned catchment |
schoolID |
Assigned school ID for the household |
num_people |
Total number of people in the household |
# Simulate catchment area catch_df <- catchment_sim(4, 5, shape = 2.5, rate = 1.3) # Simulate elementary schools using default gamma distribution elementary_df <- elementary_pop(catch_df, shape = 3.3, rate = 0.015) # Simulate households with children house_children <- subpop_children(elementary_df, n = 2, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.6) # Using custom distributions house_children2 <- subpop_children(elementary_df, n = 3, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.6, parent_dist = stats::rnorm, mean = 0.5, sd = 0.1, child_dist = stats::rbeta, shape1 = 2, shape2 = 2, age_dist = stats::runif)# Simulate catchment area catch_df <- catchment_sim(4, 5, shape = 2.5, rate = 1.3) # Simulate elementary schools using default gamma distribution elementary_df <- elementary_pop(catch_df, shape = 3.3, rate = 0.015) # Simulate households with children house_children <- subpop_children(elementary_df, n = 2, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.6) # Using custom distributions house_children2 <- subpop_children(elementary_df, n = 3, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.6, parent_dist = stats::rnorm, mean = 0.5, sd = 0.1, child_dist = stats::rbeta, shape1 = 2, shape2 = 2, age_dist = stats::runif)
Simulation of households without children using data frames from elementary_pop() and subpop_children() functions.
subpop_noChildren( df, df2, prop_house_size = NULL, prop_house_Children = NULL, house_size = NULL )subpop_noChildren( df, df2, prop_house_size = NULL, prop_house_Children = NULL, house_size = NULL )
df |
simulated output data frame from subpop_children function |
df2 |
simulated output data frame from elementary_pop function |
prop_house_size |
vector of proportions for households with 1, 2, 3, 4, 5+ members (optional) |
prop_house_Children |
proportion of households with children (optional) |
house_size |
vector of random numbers for household size simulation (optional) |
This function can be used interactively or with pre-specified parameters. If proportions are not provided, the user will be prompted to enter them. The function calculates the number of households without children for each catchment area based on the proportion of households with children.
A data frame representing the simulated population of households without children, including:
catchID |
Catchment area ID |
houseID |
Unique identifier for each household |
num_people |
Number of people in the household |
schoolPop |
Total population of elementary school assigned for the household |
xStart |
Starting X-coordindate for assigned catchment |
xEnd |
End X-coordindate for assigned catchment |
yStart |
Starting Y-coordindate for assigned catchment |
yEnd |
End Y-coordindate for assigned catchment |
# Simulate catchment areas catch_df <- catchment_sim(4, 5, shape = 2.5, rate = 1.8) # Simulate elementary schools using default gamma distribution elementary_df <- elementary_pop(catch_df, shape = 3.1, rate = 0.015) # Simulate households with children house_children <- subpop_children(elementary_df, n = 3, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.6) # Simulate households without children using pre-specified proportions house_noChild <- subpop_noChildren(house_children, elementary_df, prop_house_size = c(0.2, 0.3, 0.25, 0.15, 0.1), prop_house_Children = 0.3)# Simulate catchment areas catch_df <- catchment_sim(4, 5, shape = 2.5, rate = 1.8) # Simulate elementary schools using default gamma distribution elementary_df <- elementary_pop(catch_df, shape = 3.1, rate = 0.015) # Simulate households with children house_children <- subpop_children(elementary_df, n = 3, prop_parent_couple = 0.7, prop_children_couple = c(0.3, 0.5, 0.2), prop_children_lone = c(0.4, 0.4, 0.2), prop_elem_age = 0.6) # Simulate households without children using pre-specified proportions house_noChild <- subpop_noChildren(house_children, elementary_df, prop_house_size = c(0.2, 0.3, 0.25, 0.15, 0.1), prop_house_Children = 0.3)
This function provides a summary of the alarm metrics, including mean and variance for each metric, the best lag and threshold values, and lists of reference dates and best prediction dates.
## S3 method for class 'alarm_metrics_summary' summary(object, ...)## S3 method for class 'alarm_metrics_summary' summary(object, ...)
object |
A list of class |
... |
Additional arguments passed to the function (currently unused). |
A printed summary of the alarm metrics.
set.seed(123) # Generate sample metrics generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix() ) metrics <- structure(sample_metrics, class = c("alarm_metrics", "list")) # Generate sample best models generate_best_model <- function() { data.frame( ScYr = rep(1:3, each = 100), Date = rep(1:100, 3), Alarm = sample(c(0, 1), 300, replace = TRUE, prob = c(0.9, 0.1)), lag = sample(1:15, 300, replace = TRUE), thres = runif(300, 0.1, 0.6) ) } best_models <- list( best.FAR = generate_best_model(), best.ADD = generate_best_model(), best.AATQ = generate_best_model(), best.FATQ = generate_best_model(), best.WAATQ = generate_best_model(), best.WFATQ = generate_best_model() ) # Generate sample epidemic data epidemic_data <- data.frame( ScYr = rep(1:3, each = 365), Date = rep(1:365, 3), ref_date = c(rep(0, 364), 1) ) # Create alarm metrics summary data <- create_alarm_metrics_summary(metrics, best_models, epidemic_data) # Print summary summary(data)set.seed(123) # Generate sample metrics generate_metric_matrix <- function() { matrix(runif(15 * 11), nrow = 15, ncol = 11, dimnames = list(paste("Lag", 1:15), paste("Threshold", seq(0.1, 0.6, by = 0.05)))) } sample_metrics <- list( FAR = generate_metric_matrix(), ADD = generate_metric_matrix(), AATQ = generate_metric_matrix(), FATQ = generate_metric_matrix(), WAATQ = generate_metric_matrix(), WFATQ = generate_metric_matrix() ) metrics <- structure(sample_metrics, class = c("alarm_metrics", "list")) # Generate sample best models generate_best_model <- function() { data.frame( ScYr = rep(1:3, each = 100), Date = rep(1:100, 3), Alarm = sample(c(0, 1), 300, replace = TRUE, prob = c(0.9, 0.1)), lag = sample(1:15, 300, replace = TRUE), thres = runif(300, 0.1, 0.6) ) } best_models <- list( best.FAR = generate_best_model(), best.ADD = generate_best_model(), best.AATQ = generate_best_model(), best.FATQ = generate_best_model(), best.WAATQ = generate_best_model(), best.WFATQ = generate_best_model() ) # Generate sample epidemic data epidemic_data <- data.frame( ScYr = rep(1:3, each = 365), Date = rep(1:365, 3), ref_date = c(rep(0, 364), 1) ) # Create alarm metrics summary data <- create_alarm_metrics_summary(metrics, best_models, epidemic_data) # Print summary summary(data)
This function provides a summary of SSIR epidemic simulation. It calculates and displays the statistics.
## S3 method for class 'ssir_epidemic' summary(object, ...)## S3 method for class 'ssir_epidemic' summary(object, ...)
object |
An object of class "ssir_epidemic", typically the result of calling ssir() with rep = 1 or NULL. |
... |
Additional arguments affecting the summary produced. |
No return value, called for side effects.
# Run multiple simulations result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) # Display summary summary(result)# Run multiple simulations result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7) # Display summary summary(result)
This function provides a summary of multiple SSIR epidemic simulations. It calculates and displays average statistics across all simulations.
## S3 method for class 'ssir_epidemic_multi' summary(object, ...)## S3 method for class 'ssir_epidemic_multi' summary(object, ...)
object |
An object of class "ssir_epidemic_multi", typically the result of calling ssir() with rep > 1. |
... |
Additional arguments affecting the summary produced. |
No return value, called for side effects.
# Run multiple simulations multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 100) # Display summary summary(multi_result)# Run multiple simulations multi_result <- ssir(N = 10000, T = 300, alpha = 0.3, inf_period = 4, inf_init = 32, report = 0.02, lag = 7, rep = 100) # Display summary summary(multi_result)