def generate_csv_files(args, data_handler):
"""
Generate the CSV file used to call the R script.
:param Namespace args:
:param DataHandler data_handler:
"""
assert isinstance(args, argparse.Namespace)
assert isinstance(data_handler, DataHandler)
biomarkers = data_handler.get_biomarker_names()
measurements = data_handler.get_measurements_as_dict(min_visits=args.min_visits,
select_training_set=True,
exclude_deceased=args.exclude_deceased)
for biomarker in biomarkers:
print log.INFO, 'Generating output CSV for {0}...'.format(biomarker)
samples_file = data_handler.get_samples_file(biomarker)
writer = csv.writer(open(samples_file, 'wb'), delimiter=',')
writer.writerow(['rid', 'progress', 'value', 'diagnosis'])
subjects = set()
num_samples = 0
for rid, visits in measurements.items():
for _, visit_data in visits.items():
try:
progress = DataHandler.safe_cast(visit_data['progress'], int)
value = DataHandler.safe_cast(visit_data[biomarker], float)
diagnosis = DataHandler.safe_cast(visit_data['DX.scan'], float)
if progress is not None and value is not None:
writer.writerow([rid, progress, value, diagnosis])
subjects.add(rid)
num_samples += 1
except KeyError:
pass
print log.RESULT, 'Collected {0} samples from {1} subjects.'.format(num_samples, len(subjects))
def print_training_samples_statistics(args, data_handler):
biomarkers = data_handler.get_biomarker_names()
measurements = data_handler.get_measurements_as_dict(min_visits=args.min_visits,
select_training_set=True,
no_regression=True)
for biomarker in biomarkers:
subjects = set()
num_samples = 0
for rid, visits in measurements.items():
for _, visit_data in visits.items():
try:
progress = DataHandler.safe_cast(visit_data['progress'], int)
value = DataHandler.safe_cast(visit_data[biomarker], float)
if progress is not None and value is not None:
subjects.add(rid)
num_samples += 1
except KeyError:
pass
print log.RESULT, 'Biomarker {0}: collected {1} samples from {2} subjects.'.format(biomarker, num_samples, len(subjects))
def print_training_samples_statistics(args, data_handler):
biomarkers = data_handler.get_biomarker_names()
measurements = data_handler.get_measurements_as_dict(
min_visits=args.min_visits,
select_training_set=True,
no_regression=True)
for biomarker in biomarkers:
subjects = set()
num_samples = 0
for rid, visits in measurements.items():
for _, visit_data in visits.items():
try:
progress = DataHandler.safe_cast(visit_data['progress'],
int)
value = DataHandler.safe_cast(visit_data[biomarker], float)
if progress is not None and value is not None:
subjects.add(rid)
num_samples += 1
except KeyError:
pass
print log.RESULT, 'Biomarker {0}: collected {1} samples from {2} subjects.'.format(
biomarker, num_samples, len(subjects))
