fitAlgs.fitSims module

Author:Dominic Hunt

exception fitAlgs.fitSims.ActionError

Bases: Exception

class fitAlgs.fitSims.FitSim(participant_choice_property='Actions', participant_reward_property='Rewards', model_fitting_variable='ActionProb', task_stimuli_property=None, fit_subset=None, action_options_property=None, float_error_response_value=1e-100)

Bases: object

A class for fitting data by passing the participant data through the model.

This has been setup for fitting action-response models

Parameters:

• participant_choice_property (string, optional) – The participant data key of their action choices. Default 'Actions'
• participant_reward_property (string, optional) – The participant data key of the participant reward data. Default 'Rewards'
• model_fitting_variable (string, optional) – The key to be compared in the model data. Default 'ActionProb'
• task_stimuli_property (list of strings or None, optional) – The keys containing the stimuli seen by the participant before taking a decision on an action. Default None
• action_options_property (string or None or list of ints, optional) – The name of the key in partData where the list of valid actions can be found. If None then the action list is considered to stay constant. If a list then the list will be taken as the list of actions that can be taken at each instance. Default None
• float_error_response_value (float, optional) – If a floating point error occurs when running a fit the fitter function will return a value for each element of fpRespVal. Default is 1/1e100
• fit_subset (float('Nan'), None, "rewarded", "unrewarded", "all" or list of int, optional) – Describes which, if any, subset of trials will be used to evaluate the performance of the model. This can either be described as a list of trial numbers or, by passing - "all" for fitting all trials - float('Nan') or "unrewarded" for all those trials whose feedback was float('Nan') - "rewarded" for those who had feedback that was not float('Nan') Default None, which means all trials will be used.

Name

The name of the fitting type

Type:string

See also

fitAlgs.fitAlg.FitAlg

The general fitting class

find_name()

Returns the name of the class

fitness(*model_parameters)

Used by a fitter to generate the list of values characterising how well the model parameters describe the participants actions.

Parameters:model_parameters (list of floats) – A list of the parameters used by the model in the order previously defined Returns:model_performance – The choices made by the model that will be used to characterise the quality of the fit. Return type:list of floats

See also

fitAlgs.fitSims.FitSim.participant()

Fits participant data

fitAlgs.fitAlg.fitAlg()

The general fitting class

fitAlgs.fitAlg.fitAlg.fitness()

The function that this one is called by

fitted_model(*model_parameters)

Simulating a model run with specific parameter values

Parameters:*model_parameters (floats) – The model parameters provided in the order defined in the model setup Returns:model_instance Return type:model.modelTemplate.Model class instance

get_model_parameters(*model_parameters)

Compiles the model parameter arguments based on the model parameters

Parameters:model_parameters (list of floats) – The parameter values in the order extracted from the modelSetup parameter dictionary Returns:parameters – The kwarg model parameter arguments Return type:dict

get_model_properties(*model_parameters)

Compiles the kwarg model arguments based on the model_parameters and previously specified other parameters

Parameters:model_parameters (list of floats) – The parameter values in the order extracted from the modelSetup parameter dictionary Returns:model_properties – The kwarg model arguments Return type:dict

info()

The dictionary describing the fitters algorithm chosen

Returns:fitInfo – The dictionary of fitters class information Return type:dict

static participant_sequence_generation(participant_data, choice_property, reward_property, stimuli_property, action_options_property)

Finds the stimuli in the participant data and returns formatted observations

Parameters:

• participant_data (dict) – The participant data
• choice_property (string) – The participant data key of their action choices.
• reward_property (string) – The participant data key of the participant reward data
• stimuli_property (string or None or list of strings) – A list of the keys in partData representing participant stimuli
• action_options_property (string or None or list of strings, ints or None) – The name of the key in partData where the list of valid actions can be found. If None then the action list is considered to stay constant. If a list then the list will be taken as the list of actions that can be taken at every trialstep. If the list is shorter than the number of trialsteps, then it will be considered to be a list of valid actions for each trialstep.

Returns:

participant_sequence – Each list element contains the observation, action and feedback for each trial taken by the participant

Return type:

list of three element tuples

prepare_sim(model, model_parameters, model_properties, participant_data)

Set up the simulation of a model following the behaviour of a participant

Parameters:

• model (model.modelTemplate.Model inherited class) – The model you wish to try and fit values to
• model_parameters (dict) – The model initial parameters
• model_properties (dict) – The model static properties
• participant_data (dict) – The participant data

Returns:

Return type:

fitness

exception fitAlgs.fitSims.FitSubsetError

Bases: Exception

exception fitAlgs.fitSims.StimuliError

Bases: Exception