importDadiResults

About

An R function to extract data from multiple ∂a∂i (aka dadi, Diffusion Approximations for Demographic Inference) output files. Parses, cleans, and aggregates dadi results, calculates statistical metrics, and outputs a formatted table of comparative results.

Features

• Imports multiple dadi output files.
• Extracts and parses data,
• Handles broken rows and unwanted characters.
• Identifies and extract parameters based on a given parameter list.
• Computes AIC and AICc values for model comparison.
• Outputs a data frame with processed results.

Requires dplyr. Built with R 4.3

Inputs

• filenames (character vector): directory of dadi results files.
• param.list (list): models and parameter names
• n.list (list): sample sizes for each dataset (required for AICc calculation)
• verbose (boolean): print detailed information during processing if TRUE. Default is FALSE.

Outputs

A data.frame with a row for each dadi model, and the following columns:

• start: initial value for each run
• posterior: estimates for parameters
• loglikelihood: log-likelihood value
• theta: Estimated theta (4Ne​μ)
• prior: parameter priors
• upper: Estiamted upper bounds
• lower: Estimated lower bounds
• pts: number of grid points used
• flnms: File names
• AIC: Akaike Information Criterion
• AICc: AIC corrected for small sample sizes.

Citations

Gutenkunst, R., Hernandez, R., Williamson, S. et al. Diffusion Approximations for Demographic Inference: DaDi. Nat Prec (2010). 10.1038/npre.2010.4594.1