help=
"If provided with a /path/to/save/file.npz, the script saves the results of the analysis to file."
)
args = parser.parse_args()
# load spreadsheet / data frame
datasets = load_dataframe(args.spreadsheet_file_path)
# check contents of spreadsheet
check_dataframe(datasets,
columns=[
'file_path_raw_signals',
'file_path_preprocessed_signals',
'file_path_manual_state_annotation',
],
column_to_dtype={
'file_path_raw_signals': str,
'file_path_preprocessed_signals': str,
'file_path_manual_state_annotation': str,
})
if args.only:
datasets = datasets.loc[np.in1d(range(len(datasets)), args.only)]
# --------------------------------------------------------------------------------
print("Loading data sets...")
signal_arrays = []
state_vectors = []
for ii, dataset in datasets.iterrows():
"--model",
help=
"Use pre-trained model saved at /path/to/trained_model.pickle. If none is provided, the test is run in a hold-one-out fashion instead."
)
args = parser.parse_args()
# load spreadsheet / data frame
datasets = load_dataframe(args.spreadsheet_file_path)
# check contents of spreadsheet
check_dataframe(datasets,
columns=[
'file_path_preprocessed_signals',
'file_path_manual_state_annotation',
],
column_to_dtype={
'file_path_preprocessed_signals': str,
'file_path_manual_state_annotation': str,
})
if args.only:
datasets = datasets.loc[np.in1d(range(len(datasets)), args.only)]
# --------------------------------------------------------------------------------
print("Loading data sets...")
signal_arrays = []
state_vectors = []
for ii, dataset in datasets.iterrows():
print("{} ({}/{})".format(dataset['file_path_preprocessed_signals'],
)
args = parser.parse_args()
# load spreadsheet / data frame
datasets = load_dataframe(args.spreadsheet_file_path)
# check contents of spreadsheet
check_dataframe(
datasets,
columns=[
'file_path_raw_signals',
'sampling_frequency_in_hz',
'file_path_{}_state_annotation'.format(args.annotation_type_a),
'file_path_{}_state_annotation'.format(args.annotation_type_b),
] + state_annotation_signals,
column_to_dtype={
'file_path_raw_signals':
str,
'sampling_frequency_in_hz':
(int, float, np.int, np.float, np.int64, np.float64),
'file_path_{}_state_annotation'.format(args.annotation_type_a):
str,
'file_path_{}_state_annotation'.format(args.annotation_type_b):
str,
})
if args.only:
datasets = datasets.loc[np.in1d(range(len(datasets)), args.only)]
# turn off interactive mode if on
plt.ioff()
type = int,
help = 'Indices corresponding to the rows to use (default: all). Indexing starts at zero.'
)
args = parser.parse_args()
# load spreadsheet / data frame
datasets = load_dataframe(args.spreadsheet_file_path)
# check contents of spreadsheet
check_dataframe(datasets,
columns = [
'file_path_preprocessed_signals',
'file_path_automated_state_annotation',
'file_path_review_intervals',
],
column_to_dtype = {
'file_path_preprocessed_signals' : str,
'file_path_automated_state_annotation' : str,
'file_path_review_intervals' : str,
}
)
if args.only:
datasets = datasets.loc[np.in1d(range(len(datasets)), args.only)]
annotator = StateAnnotator()
annotator.load(args.trained_model_file_path)
for ii, dataset in datasets.iterrows():
print("{} ({}/{})".format(dataset['file_path_preprocessed_signals'], ii+1, len(datasets)))
nargs='+',
type=int,
help=
'Indices corresponding to the rows to use (default: all). Indexing starts at zero.'
)
args = parser.parse_args()
# load spreadsheet / data frame
datasets = load_dataframe(args.spreadsheet_file_path)
# check contents of spreadsheet
check_dataframe(datasets,
columns=[
'file_path_sleepsign_state_annotation',
'file_path_manual_state_annotation',
],
column_to_dtype={
'file_path_sleepsign_state_annotation': str,
'file_path_manual_state_annotation': str,
})
if args.only:
datasets = datasets.loc[np.in1d(range(len(datasets)), args.only)]
for ii, dataset in datasets.iterrows():
print("{} ({}/{})".format(
dataset['file_path_sleepsign_state_annotation'], ii + 1,
len(datasets)))
old_file_path = dataset['file_path_sleepsign_state_annotation']
new_file_path = dataset['file_path_manual_state_annotation']
convert_sleepsign_hypnogram(old_file_path, new_file_path)
parser.add_argument('--only',
nargs = '+',
type = int,
help = 'Indices corresponding to the rows to use (default: all). Indexing starts at zero.'
)
args = parser.parse_args()
# load spreadsheet / data frame
datasets = load_dataframe(args.spreadsheet_file_path)
# check contents of spreadsheet
check_dataframe(datasets,
columns = [
'file_path_{}_state_annotation'.format(args.annotation_type),
'file_path_{}_state_annotation_mat'.format(args.annotation_type)
],
column_to_dtype = {
'file_path_{}_state_annotation'.format(args.annotation_type) : str,
'file_path_{}_state_annotation_mat'.format(args.annotation_type) : str,
}
)
if args.only:
datasets = datasets.loc[np.in1d(range(len(datasets)), args.only)]
for ii, dataset in datasets.iterrows():
print("{} ({}/{})".format(dataset['file_path_{}_state_annotation'.format(args.annotation_type)], ii+1, len(datasets)))
old_file_path = dataset['file_path_{}_state_annotation'.format(args.annotation_type)]
new_file_path = dataset['file_path_{}_state_annotation_mat'.format(args.annotation_type)]
convert_hypnogram_to_mat(old_file_path, new_file_path, mapping=state_to_int, time_resolution=1)