ToEstimate
ToEstimate(params_dict)

Input parameters: A dictionary with the following potential entries

• params: The name of the parameter being fitted (Symbol)
• indices: The index of the parameter being fitted (Integer or Tuple{Integer,Integer})
• factor: Factor to multiply parameter being fitted to have it in the correct units (Float64)
• symmetric: For PairParam, if the parameter is symmetric or asymmetric (Bool)
• cross_assoc: For AssocParam, if the parameter is for cross-association (Bool)
• recombine: For PairParam, if the combining rules must be applied for unlike interactions (Bool)
• lower: Lower bound for the parameter (Float64)
• upper: Upper bound for the parameter (Float64)
• guess: Initial guess for the parameter (Float64)

Output:

A ToEstimate struct

Description

Turns the input parameter dictionary into a ToEstimate struct to be used within the parameter estimation.

Estimation
Estimation(model::EoSModel,toestimate::Dict,filepaths;ignorefield = Vector{String},objective_form = mse(pred,exp) = ((pred-exp)/exp)^2)

Input parameters:

• model: The initial model containing the species we wish to parameterise
• toestimate: The dictionary of parameters being fitted
• filepaths or filepaths_weights: The location of the data files used to fit. Can also contain the weights of each dataset
• ignorefield: Specify which EoSModel fields to ignore in the main model
• objective_form: Specify the functional form of the objective function in the form objective_form(pred,exp)

Output:

Estimator object which contains the following:

• model: The model whose parameters will be varied
• initial_model: The initial model before parameterisation
• toestimate: ToEstimate struct which contains all the information on the parameters
• data: Vector of EstimationData structs where all the information on the data is stored
• ignorefield: Vector of fields to ignore in the parameter estimation
• objective_form: Function to evaluate the error measure for the objective function

The following objects are also output:

• objective: The objective function which is used to fit the parameters
• x0: Initial guesses for the parameters
• upper: Upper bounds for the parameters
• lower: Lower bounds for the parameters

Description

Produces the estimator and other useful objects used within parameter estimation

EstimationData
EstimationData(filepaths)

Input parameters:

• filepaths or filepaths_weights: The filepath of the data used in parameter estimation. Optionally, a tuple containing the weights of each dataset.

Output:

An EstimationData struct with the following fields:

• method: The property estimation method which is used to obtain predictions for a given input
• inputs_name: The variable names for the inputs
• outputs_name: The variable names for the outputs
• inputs: Vector for each input
• outputs: Vector for each output
• weights: The weight for this particular dataset

Description

For a given input data set, produce an EstimationData struct.

return_model
return_model(estimation,model,params)

Input parameters:

• estimation: The estimator object
• model: The model whose parameters we are varying
• params: The new parameters which we want to change

Output:

• model: The new model with the updated parameters

Description

Based on the parameters provided and the estimator, a new model is produced from the input.

objective_function
objective_function(estimation,params)

Input parameters:

• estimation: The estimator object
• params: The new parameters which we want to evaluate the objective function for

Output: The relate root mean square error given the data and parameters provided

Description

The objective function used within parameter estimation.