cxr_pm_bootstrap: Standard error estimates for model parameters

In RadicalCommEcol/cxr: A Toolbox for Modelling Species Coexistence in R

Standard error estimates for model parameters

Description

Computes bootstrap standard errors for a given population dynamics model. This function is provided for completeness, but error calculation is integrated in the function cxr_pm_fit.

Usage

cxr_pm_bootstrap(
  fitness_model,
  optimization_method,
  data,
  focal_column,
  covariates,
  init_par,
  lower_bounds,
  upper_bounds,
  fixed_parameters,
  bootstrap_samples
)

Arguments

fitness_model	function returning a single value to minimize, given a set of parameters and a fitness metric
optimization_method	numerical optimization method
data	dataframe with observations in rows and two sets of columns: fitness: fitness metric for the focal individual neighbours: columns with user-defined names with number of neighbours for each group
focal_column	optional integer value giving the position, or name, of the column with neighbours from the same species as the focal one. This is necessary if "alpha_intra" is specified.
covariates	optional matrix with observations in rows and covariates in columns. Each cell is the value of a covariate in a given observation.
init_par	1d vector of initial parameters
lower_bounds	1d vector of lower bounds
upper_bounds	1d vector of upper bounds
fixed_parameters	optional list specifying values of fixed parameters, with components "lambda","alpha_intra","alpha_inter","lambda_cov", and "alpha_cov".
bootstrap_samples	how many bootstrap samples to compute.

Value

1d vector, the standard error of each parameter in init_par

RadicalCommEcol/cxr documentation built on Oct. 29, 2023, 10:07 p.m.