parameters - SKILL.md Agent Skill

name: parameters description: Skill for retrieving and managing epidemiological parameters (incubation periods, serial intervals) using epiparameter.

Epiverse-TRACE Parameters

[!IMPORTANT] Use the Introspection Protocol: See epiverse-overview skill for the protocol. Before generating code, verify package APIs and functions using R introspection commands.

[!NOTE] Workflow Context: This skill's code should be included in Quarto document code chunks (not separate R scripts). See the reporting skill for the complete workflow structure.

This skill provides access to a library of epidemiological parameters from the literature and tools for parameter manipulation.

Packages

`epiparameter`

Purpose: Classes and helper functions for working with epidemiological parameters.

Key Functions:

epiparameter_db(): Load parameters from library (disease, epi_name, author filters)
epiparameter(): Create custom parameter object
parameter_tbl(): View parameters as table
plot(): Visualize parameter distribution
convert_params_to_summary_stats(): Convert distribution parameters to mean/SD
convert_summary_stats_to_params(): Convert mean/SD to distribution parameters
extract_param(): Extract parameters from summary statistics

Supported Distributions: gamma, lognormal, Weibull, normal, negative binomial, geometric

`epiparameterDB`

Purpose: Database of epidemiological parameters extracted from literature.

Access: Via epiparameter::epiparameter_db() (recommended) or direct JSON access

Coverage: 23 diseases, 125+ parameter sets including incubation periods, serial intervals, generation times, onset-to-hospitalization, onset-to-death

Online Database: View at https://epiverse-trace.github.io/epiparameter/articles/database.html

Typical Workflow

Finding Parameters

# Search by disease and parameter type
param <- epiparameter_db(
  disease = "Ebola Virus Disease",
  epi_name = "incubation period",
  single_epiparameter = TRUE
)

# View all available parameters
all_params <- epiparameter_db()
parameter_tbl(all_params)

Using Parameters

# Extract distribution function
dist_func <- function(x) dgamma(x, shape = param$shape, scale = param$scale)

# Use in analysis
cfr_static(data, delay_density = dist_func)

Creating Custom Parameters

# When parameter not in library
custom_param <- epiparameter(
  disease = "COVID-19",
  epi_name = "onset to hospitalization",
  prob_distribution = create_prob_distribution(
    prob_distribution = "lnorm",
    prob_distribution_params = c(meanlog = 1.5, sdlog = 0.5)
  )
)

Parameter Conversion

# Convert summary statistics to distribution parameters
params <- convert_summary_stats_to_params(
  mean = 10,
  sd = 5,
  distribution = "gamma"
)

Common Parameter Types

Incubation period: Time from infection to symptom onset
Serial interval: Time between symptom onset in infector and infectee
Generation time: Time between infection in infector and infectee
Onset to hospitalization: Time from symptoms to hospital admission
Onset to death: Time from symptoms to death
Hospitalisation to death: Time from admission to death
Offspring distribution: Number of secondary infections per case

Best Practices

Always cite the source of parameters (use get_citation())
Check parameter applicability (population, setting, time period)
Use single_epiparameter = TRUE when expecting one result
Visualize distributions with plot() to verify plausibility
Document parameter choices in analysis reports
Consider uncertainty in parameter estimates

Contributing Parameters

To add parameters to the library:

Use the public Google Sheet
Or submit PR to epiparameterDB

Integration with Other Skills

simulation: Use parameters to simulate realistic outbreaks
analysis: Use delay distributions for CFR correction
reporting: Document parameter sources and justifications