compute_structure_SBM.py

#!/usr/bin/env python3
import sys
import os
import argparse
import logging
from src.file_handler import file_handler
import src.graph_tool_functions as gt
from src.version import repo_version
__version__ = repo_version()



DESCRIPTION = '''
Identify large-scale network structure by fitting the stochastic block model, by minimizing its description length using an agglomerative heuristic.

For more info:
https://graph-tool.skewed.de/static/doc/demos/inference/inference.html
https://graph-tool.skewed.de/static/doc/autosummary/graph_tool.inference.minimize_blockmodel_dl.html
'''



## Pass arguments.
def main():
	## Pass command line arguments. 
	parser = argparse.ArgumentParser(formatter_class=argparse.RawDescriptionHelpFormatter, description=DESCRIPTION)
	parser.add_argument('-i', '--input_graph', metavar='graph.gt.xz', 
		required=False, default='/dev/stdin', type=lambda x: file_handler(x, 'rs'), 
		help='Input graph file (default: stdin)'
	)
	parser.add_argument('-o', '--output_graph', metavar='graph.SBM.gt.xz',
		required=False, default='/dev/stdout', type=lambda x: file_handler(x, 'ws'),
		help='Output graph file with fitted stochastic block model (default: stdout)'
	)
	parser.add_argument('-n', '--ncpus',
		required=False, default=6, type=int,
		help='Number of CPUs to use (default: %(default)s)'
	)
	parser.add_argument('--version', action='version',
		version=f"{__version__}",
		help='Print repo version'
	)
	parser.add_argument('--debug', 
		required=False, action='store_true', 
		help='Print DEBUG info (default: %(default)s)'
	)
	args = parser.parse_args()
	
	## Set up basic debugger
	if args.debug:
		logging.getLogger().setLevel(logging.DEBUG)
	
	logging.debug('%s', args) ## DEBUG
	logging.info('Version: %s', __version__) ##INFO
	
	## Workflow
	with args.input_graph as in_graph_file, args.output_graph as out_graph_file:
		gt.set_ncpus(args.ncpus)
		G = gt.load_graph_file(in_graph_file)
		G = gt.run_minimize_blockmodel_dl(G)
		gt.save_graph_file(G, out_graph_file)



if __name__ == '__main__':
	log_format = "%(asctime)s [%(levelname)s] "+os.path.basename(__file__)+": %(message)s"
	logging.basicConfig(format=log_format, stream=sys.stderr, level=logging.INFO)
	main()