if __name__ == '__main__':
parser = argparse.ArgumentParser(
description='Generate D3M dataset config files')
parser.add_argument('-a', '--absolute', action='store_true')
parser.add_argument('-b', '--base-dir', default='data', nargs='?')
parser.add_argument('-d', '--datasets', default='datasets', nargs='?')
parser.add_argument('-o', '--output', default='output', nargs='?')
parser.add_argument('-c', '--config-dir', required=True)
parser.add_argument('-p', '--problem', default='', nargs='?')
parser.add_argument('dataset', nargs='+')
args = parser.parse_args()
if args.absolute:
base_dir = make_abs(args.base_dir, os.getcwd())
datasets = make_abs(args.datasets, base_dir)
output = make_abs(args.output, base_dir)
else:
base_dir = args.base_dir
datasets = os.path.join(base_dir, args.datasets)
output = os.path.join(base_dir, args.output)
for dataset in args.dataset:
for phase, phase_filename in PHASES.items():
config = build_config(dataset, datasets, phase, args.problem,
output)
filename = '{}_{}.json'.format(dataset, phase_filename)
with open(os.path.join(args.config_dir, filename), 'w') as f:
json.dump(config, f, indent=4)
return pd.DataFrame(data)[STATS_COLUMNS]
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Get Dataset summary')
parser.add_argument('-i', '--input', default='data/datasets', nargs='?')
parser.add_argument('-o', '--output', nargs='?')
parser.add_argument('datasets', nargs='*')
args = parser.parse_args()
logging_setup()
args.input = make_abs(args.input, os.getcwd())
if not args.datasets:
args.datasets = os.listdir(args.input)
print("Processing Datasets: {}".format(args.datasets))
output = get_stats(args.datasets, args.input)
if args.output:
print("Storing report as {}".format(args.output))
output.to_csv(args.output, index=False)
else:
print(output)
def _path_type(string):
try:
return make_abs(string)
except ValueError:
error = "Not a valid path: '{0}'.".format(string)
raise argparse.ArgumentTypeError(error)