solution_statistics function

Calculate Solution Statistics

Attempts to determine the variance/covariance matrix, confidence intervals and CV percent for a list of parameter estimates parameters. This method was taken from the ADAPT 5 User's Guide chapter 3.

solution_statistics(parameters, cfg)

Arguments

• parameters: list of parameter estimates
• cfg: ubiquity system object

Returns

list containing information about the provided estimates

Details

The returned list has the following format:

• objective - objective function value

• num_observations - number of observations

• degrees_of_freedom - degrees of freedom

• aic - Akaike information criterion

• bic - Bayesian (Schwarz) information criterion

• covariance - variance covariance matrix

• wls - defined for weighted least squares objective with the following elements:

  • jacobian - Jacobian matrix
  • weights - diagonal matrix of weights
  • error_variance - diagonal matrix of variances

• ml - defined for maximum likelihood objective with the following elements:

  • M - Jacobian matrix with block for variance parameters

• coefficient_of_variation$pname - CV percent for parameter pname

• confidence_interval$lower_bound$pname - Lower bound of the confidence interval for pname

• confidence_interval$upper_bound$pname - Upper bound of the confidence interval for pname

See Also

Vignette on estimation (vignette("Estimation", package = "ubiquity"))