Add more default parameters to deteCT

[AnderBiguri]

Currently the default settings for deteCT are to do a sino2recon in mode2:

LION/LION/data_loaders/deteCT.py

Line 191 in b80d057

def default_parameters():

It would be interesting to have more default parameters, such as

@staticmethod
def default_denoising_parameters():
    ....
@staticmethod
def default_beam_hardening_parameters():
    ....

That automatically give you a dataset with a standard config. Note: I would not change the geometry here, this is for Experiments classes. These parameters are only to define what is loaded in (input,target) pairs.

One thus could do in their code:

from LION.data_loaders.deteCT import deteCT

dataset = deteCT('training') # this loads default parameters
dataset = deteCT('training',parameters=deteCT.default_parameters()) #equivalent to the above

dataset = deteCT('training', parameters=deteCT.default_denoising_parameters()) # example of new default setting)

This could be even extended to the Experiment class. Currently, when you define a new Experiment, at some point you do Experiment.get_dataset_parameters(dataset_name) and this returns default parameters. We could modify this functions:

LION/LION/experiments/ct_experiments.py

Line 122 in b80d057

def get_dataset_parameters(dataset, geo=None):

to return specific default parameters. Or alternatively, new defined Experiments can directly only support certain datasets, e.g. deteCT and replace Experiment.get_dataset_parameters(dataset_name) with deteCT.default_denoising_parameters() for example.