Added graph stats script

import argparse

import json

import collections

import numpy as np

import networkx as nx

parser = argparse.ArgumentParser(description="Analyze a given file.")

parser.add_argument("-general", help="Show some general info on the graph", action="store_true")

parser.add_argument("-top-nodes", type=int, metavar="N", help="Show N top nodes according to degree")

parser.add_argument("-detail", type=str, choices=["defendant","victim","participant","verdict","offence","punishment","description"], help="Show details on selected category over all trials")

parser.add_argument("file", help="Path to the graph file to analyze")

args = parser.parse_args()

inputfile = args.file

print("Processing", inputfile)

with open(inputfile) as f:

	data = json.load(f)

nodes = data["nodes"]

edges = data["links"]

G = nx.MultiDiGraph()

for edge in edges:

	source = edge["source"]

	target = edge["target"]

	G.add_edge(source, target)

if args.general:

	print("\nNumber of nodes:", len(nodes))

	print("Number of edges:", len(edges))

	print("Number of trials:", len([x for x in nodes if x['label'] == "Trial"]))

	output_dict = [x for x in nodes if 'category' in x['nodeobj'] and x['nodeobj']['category'] == 'DUMMYCAT']

	print("\nNodes with DUMMYCAT:", output_dict)

	newdict = {}

	l = [value["source"] for value in edges]

	newdict["source"] = dict(collections.Counter(l))

	l = [value["target"] for value in edges]

	newdict["target"] = dict(collections.Counter(l))

	allindegrees = []

	alloutdegrees = []

	degreedict = {}

	for key, value in newdict.items():

		if key == 'source':

			for nkey, nvalue in value.items():

				alloutdegrees.append(nvalue)

				degreedict[nkey] = nvalue

		if key == 'target':

			for nkey, nvalue in value.items():

				allindegrees.append(nvalue)

				if nkey in degreedict:

					previousval = degreedict[nkey]

					degreedict[nkey] = nvalue + previousval

				else:

					degreedict[nkey] = nvalue

	alldegrees = []

	for key, value in degreedict.items():

		alldegrees.append(value)

	print("Average indegree:", np.mean(allindegrees))

	print("Maximum indegree:", np.max(allindegrees))

	print("Average outdegree:", np.mean(alloutdegrees))

	print("Maximum outdegree:", np.max(alloutdegrees))

	print("Average degree:", np.mean(alldegrees))

	print("Maximum degree:", np.max(alldegrees))

if args.top_nodes:

	idtonode = {}

	for el in nodes:

		idtonode[el["id"]] = el

	print("\nTop " + str(args.top_nodes) + " nodes sorted by degree:")

	for el in sorted(G.degree, key=lambda x: x[1], reverse=True)[:args.top_nodes]:

		node_id = el[0]

		print("Node:", idtonode[node_id])

		print("Degree:", el[1])

if args.detail:

	trial_ids = [x['id'] for x in nodes if x['label'] == "Trial"]

	trial_to_count = {}

	total_count = 0

	if args.detail == "description":

		node_type = "offence-description"

	else:

		node_type = args.detail

	for trial in trial_ids:

		alledges = [e for e in edges if (e['source'] == trial or e['target'] == trial)]

		for edge in alledges:

			if "with-"+node_type == edge['edge_class']:

				total_count+=1

				if trial in trial_to_count:

					trial_to_count[trial] += 1

				else:

					trial_to_count[trial] = 1

	average = np.mean(list(trial_to_count.values()))

	print("\nTotal number of " + args.detail + "s:", total_count)

	node_count = 0

	category_to_count = {}

	for node in nodes:

		if  node['label'].lower() == args.detail:

			node_count += 1

			if 'category' in node['nodeobj']:

				category = node['label'] + '-' + node['nodeobj']['category']

				id = node['id']

				count = len([e for e in edges if (e['source'] == id or e['target'] == id)])

				category_to_count[category] = count

	print("Number of distinct " + args.detail + " nodes:", node_count)

	for k in category_to_count.keys():

		print(k + ": " + str(category_to_count[k]))

	print("\nAverage number of " + args.detail + "s per trial:", average)

 No newline at end of file