ml_main_cross_subject.py

from os import getcwd
from os.path import join
from os.path import dirname
from collections import OrderedDict
from numpy.random import RandomState
from hgdecode.utils import create_log
from hgdecode.utils import ml_results_saver
from hgdecode.loaders import CrossSubject
from hgdecode.classes import CrossValidation
from hgdecode.experiments import FBCSPrLDAExperiment

"""
SETTING PARAMETERS
Here you can set whatever parameter you want
"""
# setting ml_algorithm
algorithm_name = 'FBCSP_rLDA'

# setting channel_names
channel_names = ['FC5', 'FC1', 'FC2', 'FC6', 'C3', 'C4',
                 'CP5', 'CP1', 'CP2', 'CP6',
                 'FC3', 'FCz', 'FC4', 'C5', 'C1', 'C2', 'C6',
                 'CP3', 'CPz', 'CP4',
                 'FFC5h', 'FFC3h', 'FFC4h', 'FFC6h',
                 'FCC5h', 'FCC3h', 'FCC4h', 'FCC6h',
                 'CCP5h', 'CCP3h', 'CCP4h', 'CCP6h',
                 'CPP5h', 'CPP3h', 'CPP4h', 'CPP6h',
                 'FFC1h', 'FFC2h', 'FCC1h', 'FCC2h',
                 'CCP1h', 'CCP2h', 'CPP1h', 'CPP2h']

# setting data_dir & results_dir
data_dir = join(dirname(dirname(getcwd())), 'datasets', 'High-Gamma')
results_dir = join(dirname(dirname(getcwd())), 'results', 'hgdecode')

# setting name_to_start_codes
name_to_start_codes = OrderedDict([('Right Hand', [1]),
                                   ('Left Hand', [2]),
                                   ('Rest', [3]),
                                   ('Feet', [4])])

# setting random state
random_state = RandomState(1234)

# setting subject_ids
subject_ids = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14)

# %%
"""
STARTING LOADING ROUTINE & COMPUTATION
Here you can change some parameter in function calls as well
"""
# creating a log object
subj_results_dir = create_log(
    results_dir=results_dir,
    learning_type='ml',
    algorithm_or_model_name=algorithm_name,
    subject_id='subj_cross',
    output_on_file=False
)

# creating a cross-subject object for cross-subject validation
cross_obj = CrossSubject(data_dir=data_dir,
                         subject_ids=subject_ids,
                         channel_names=channel_names,
                         name_to_start_codes=name_to_start_codes,
                         resampling_freq=250,
                         train_test_split=True,
                         clean_ival_ms=(-1000, 1000),
                         epoch_ival_ms=(-1000, 1000),
                         clean_on_all_channels=False)

"""
Si potrebbe fare un soft parsing così trova le fold, poi si passa ad exp
quello che gli serve (cnt all, clean all, fold all) e si butta tutto il
resto...
"""

# creating experiment instance
exp = FBCSPrLDAExperiment(
    # signal-related inputs
    cnt=cross_obj.data,
    clean_trial_mask=cross_obj.clean_trial_mask,
    name_to_start_codes=name_to_start_codes,
    random_state=random_state,
    name_to_stop_codes=None,  # Schirrmeister: None
    epoch_ival_ms=(-1000, 1000),  # Schirrmeister: (-500, 4000)
    cross_subject_object=cross_obj,

    # bank filter-related inputs
    min_freq=[0, 10],  # Schirrmeister: [0, 10]
    max_freq=[12, 122],  # Schirrmeister: [12, 122]
    window=[6, 8],  # Schirrmeister: [6, 8]
    overlap=[3, 4],  # Schirrmeister: [3, 4]
    filt_order=3,  # filt_order: 3

    # machine learning parameters
    n_folds=14,  # Schirrmeister: ?
    n_top_bottom_csp_filters=5,  # Schirrmeister: 5
    n_selected_filterbands=None,  # Schirrmeister: None
    n_selected_features=20,  # Schirrmeister: 20
    forward_steps=2,  # Schirrmeister: 2
    backward_steps=1,  # Schirrmeister: 1
    stop_when_no_improvement=False,  # Schirrmeister: False
    shuffle=False,  # Schirrmeister: False
    average_trial_covariance=True  # Schirrmeister: True
)

# running the experiment
exp.run()

# saving results
ml_results_saver(exp=exp, subj_results_dir=subj_results_dir)

# at the very end, running cross-validation
CrossValidation.cross_validate(subj_results_dir=subj_results_dir,
                               label_names=name_to_start_codes)