wf_psf.utils.configs_handler

Configs_Handler.

A module which provides general utility methods to manage the parameters of the config files

Authors:: Jennifer Pollack <jennifer.pollack@cea.fr>

Functions

`get_run_config`(run_config_name, *config_params)	Get Run Configuration Instance.
`register_configclass`(config_class)	Register Config Class.
`set_run_config`(config_name)	Set Run Configuration Class.

Classes

`DataConfigHandler`(data_conf, ...[, ...])	DataConfigHandler.
`MetricsConfigHandler`(metrics_conf, file_handler)	MetricsConfigHandler.
`PlottingConfigHandler`(plotting_conf, ...)	PlottingConfigHandler.
`TrainingConfigHandler`(training_conf, ...)	TrainingConfigHandler.

Exceptions

ConfigParameterError([message])

Custom Config Parameter Error exception class for specific error scenarios.

exception wf_psf.utils.configs_handler.ConfigParameterError(message='An error with your config settings occurred.')[source]

Bases: Exception

Custom Config Parameter Error exception class for specific error scenarios.

class wf_psf.utils.configs_handler.DataConfigHandler(data_conf, training_model_params, batch_size=16, load_data=True)[source]

Bases: object

DataConfigHandler.

A class to handle data configuration parameters.

Parameters:

data_conf (str) – Path of the data configuration file
training_model_params (Recursive Namespace object) – Recursive Namespace object containing the training model parameters
batch_size (int) – Training hyperparameter used for batched pre-processing of data.

class wf_psf.utils.configs_handler.MetricsConfigHandler(metrics_conf, file_handler, training_conf=None)[source]

Bases: object

MetricsConfigHandler.

A class to handle metrics configuation parameters.

Parameters:

ids (tuple) – A tuple containing a string id for the Configuration Class
metrics_conf (str) – Path to the metrics configuration file
file_handler (object) – An instance of the FileIOHandler
training_conf (RecursiveNamespace object) – RecursiveNamespace object containing the training configuration parameters

Attributes:

data_conf: Get Data Conf.
metrics_conf: Get Metrics Conf.
metrics_dir: Get Metrics Directory.
plotting_conf: Get Plotting Conf.
psf_model: Get PSF Model.
training_conf: Get Training Conf.
weights_basename_filepath: Get PSF model weights filepath.
weights_path: Get Weights Path.

Methods

`call_plot_config_handler_run`(model_metrics)	Make Metrics Plots.
`run`()	Run.

call_plot_config_handler_run(model_metrics)[source]

Make Metrics Plots.

A function to call the PlottingConfigHandler run command to generate metrics plots.

Parameters:: model_metrics (dict) – A dictionary storing the metrics output generated during evaluation of the trained PSF model.

property data_conf

Get Data Conf.

A function to return an instance of the DataConfigHandler class.

Return type:: An instance of the DataConfigHandler class.

ids = ('metrics_conf',)

property metrics_conf

Get Metrics Conf.

A function to return the metrics configuration file name.

Returns:: An instance of the metrics configuration file.
Return type:: RecursiveNamespace

property metrics_dir

Get Metrics Directory.

A function that returns path of metrics directory.

Returns:: Absolute path to metrics directory
Return type:: str

property plotting_conf

Get Plotting Conf.

A function to return the plotting configuration file name.

Returns:: Name of plotting configuration file
Return type:: str

property psf_model

Get PSF Model.

A function to return an instance of the PSF model to be evaluated.

Returns:: psf_model – An instance of the PSF model to be evaluated.
Return type:: obj

run()[source]

Run.

A function to run wave-diff according to the input configuration.

property training_conf

Get Training Conf.

A function to return the training configuration file name.

Returns:: An instance of the training configuration file.
Return type:: RecursiveNamespace

property weights_basename_filepath

Get PSF model weights filepath.

A function to return the basename of the user-specified psf model weights path.

Returns:: weights_basename – The basename of the psf model weights to be loaded.
Return type:: str

property weights_path

Get Weights Path.

A function to return the full path of the user-specified psf model weights to be loaded.

Returns:: A string representing the full path to the psf model weights to be loaded.
Return type:: str

class wf_psf.utils.configs_handler.PlottingConfigHandler(plotting_conf, file_handler)[source]

Bases: object

PlottingConfigHandler.

A class to handle plotting config settings.

Parameters:

ids (tuple) – A tuple containing a string id for the Configuration Class
plotting_conf (str) – Name of plotting configuration file
file_handler (obj) – An instance of the FileIOHandler class

Methods

`check_and_update_metrics_confs`()	Check and Update Metrics Confs.
`load_metrics_into_dict`()	Load Metrics into Dictionary.
`make_dict_of_metrics`()	Make dictionary of metrics.
`run`()	Run.

check_and_update_metrics_confs()[source]

Check and Update Metrics Confs.

A function to check if user provided inputs metrics dir to add to metrics_confs dictionary.

ids = ('plotting_conf',)

load_metrics_into_dict()[source]

Load Metrics into Dictionary.

A method to load a metrics file of a trained model from a previous run into a dictionary.

Returns:: metrics_files_dict – A dictionary containing all of the metrics from the loaded metrics files.
Return type:: dict

make_dict_of_metrics()[source]

Make dictionary of metrics.

A function to create a dictionary for each metrics per run.

Returns:: A dictionary containing metrics or an empty dictionary.
Return type:: dict

run()[source]

Run.

A function to run wave-diff according to the input configuration.

class wf_psf.utils.configs_handler.TrainingConfigHandler(training_conf, file_handler)[source]

Bases: object

TrainingConfigHandler.

A class to handle training configuration parameters.

Parameters:

ids (tuple) – A tuple containing a string id for the Configuration Class
training_conf (str) – Path of the training configuration file
file_handler (object) – A instance of the FileIOHandler class

Methods

run()

Run.

ids = ('training_conf',)

run()[source]

Run.

A function to run wavediff according to the input configuration.

wf_psf.utils.configs_handler.get_run_config(run_config_name, *config_params)[source]

Get Run Configuration Instance.

A function to retrieve an instance of the appropriate configuration class for a WF-PSF run.

Parameters:

run_config_name (str) – Name of the run configuraton
*config_params (str) – Run configuration parameters used for class instantiation.

Returns:

config_class – A class instance of the selected configuration class.

Return type:

object

wf_psf.utils.configs_handler.register_configclass(config_class)[source]

Register Config Class.

A wrapper function to register all config classes in a dictionary.

Parameters:: config_class (type) – Config Class
Returns:: config_class – Config class
Return type:: type

wf_psf.utils.configs_handler.set_run_config(config_name)[source]

Set Run Configuration Class.

A function to retrieve the appropriate configuration class based on the provided config name.

Parameters:: config_name (str) – Name of config
Returns:: config_class – Name of config class
Return type:: class