Skip to contents

There is an accessor function for each slot in the S4 class FIMSFit, where the function is named get_*() and the star can be replaced with the slot name, e.g., get_input(). These accessor functions are the preferred way to access objects stored in the available slots.

Usage

get_input(x)

# S4 method for class 'FIMSFit'
get_input(x)

get_report(x)

# S4 method for class 'FIMSFit'
get_report(x)

get_obj(x)

# S4 method for class 'FIMSFit'
get_obj(x)

get_opt(x)

# S4 method for class 'FIMSFit'
get_opt(x)

get_max_gradient(x)

# S4 method for class 'FIMSFit'
get_max_gradient(x)

get_sdreport(x)

# S4 method for class 'FIMSFit'
get_sdreport(x)

get_estimates(x)

# S4 method for class 'FIMSFit'
get_estimates(x)

get_fits(x)

# S4 method for class 'FIMSFit'
get_fits(x)

get_number_of_parameters(x)

# S4 method for class 'FIMSFit'
get_number_of_parameters(x)

get_timing(x)

# S4 method for class 'FIMSFit'
get_timing(x)

get_version(x)

# S4 method for class 'FIMSFit'
get_version(x)

Arguments

x

Output returned from fit_fims().

Value

get_input() returns the list that was used to fit the FIMS model, which is the returned object from create_default_parameters().

get_report() returns the TMB report, where anything that is flagged as reportable in the C++ code is returned.

get_obj() returns the output from TMB::MakeADFun().

get_opt() returns the output from nlminb(), which is the minimizer used in fit_fims().

get_max_gradient() returns the maximum gradient found when optimizing the model.

get_sdreport() returns the list from TMB::sdreport().

get_estimates() returns a tibble of parameter values and their uncertainties from a fitted model.

get_fits() returns a tibble of input data and corresponding expected values, likelihoods, and additional metadata from a fitted model. The tibble includes the following variables:

  • module_name: Character string that describes the name of the module, e.g., "Data".

  • module_id: Integer that provides identifier for linking outputs.

  • label: Character string that describes type of data.

  • data_id: Not yet implemented/NA: Integer that will provide unique identifier for data.

  • fleet_name: Not yet implemented/NA: Character string that will provide fleet name corresponding to name provided via FIMSFrame.

  • unit: Not yet implemented/NA: Character string that will describe appropriate units for the data inputs and expected values.

  • uncertainty: Not yet implemented/NA: Character string that will describe the uncertainty specified for the data input value.

  • age: Not yet implemented/NA: Integer that will provide the age affiliated with a data input, where appropriate.

  • length: Not yet implemented/NA: Integer that will provide the length affiliated with a data input, where appropriate.

  • datestart: Not yet implemented/NA: Character string that will provide the start date for the data input, corresponding to the value provided in the input data.

  • dateend: Not yet implemented/NA: Character string that will provide the end date for the data input, corresponding to the value provided in the input data.

  • year: Not yet implemented/NA: Integer that will provide model year for the data input.

  • init: Numeric that provides the initial value for the data input.

  • expected: Numeric that provides the expected value for the data input. *NOTE: units for provided init and expected values need to be standardized.

  • log_like: Numeric that provides log-likelihood for expected value.

  • distribution: Character string that indicates the distribution used for the log-likelihood estimation.

  • re_estimated: Logical that indicates whether any random effects were estimated during model fitting. Log-likelihood values should not be directly compared between models with and without estimation of random effects.

  • log_like_cv: Not yet implemented/NA: Numeric that will indicate corresponding uncertainty for the log-likelihood value.

  • weight: Numeric that indicates data weighting applied to each data value; manually fixed at 1. *NOTE: Will need to be made responsive to user-specified or user-estimated data weighting once data weighting is added to FIMS as a feature.

get_number_of_parameters() returns a vector of integers specifying the number of fixed-effect parameters and the number of random-effect parameters in the model.

get_timing() returns the amount of time it took to run the model in seconds as a difftime object.

get_version() returns the package_version of FIMS that was used to fit the model.