Spaces:
Runtime error
Runtime error
File size: 12,753 Bytes
8044721 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 |
# GraphaLogue Analyzer
# Marco Kuhlmann <[email protected]>
import itertools
import statistics
import sys
from graph import Graph
from treewidth import quickbb
class DepthFirstSearch(object):
def __init__(self, graph, undirected=False):
self._graph = graph
self._undirected = undirected
self._enter = dict()
self._leave = dict()
self.n_runs = 0
def compute_timestamps(node, timestamp):
self._enter[node] = next(timestamp)
for edge in self._graph.find_node(node).outgoing_edges:
if not edge.tgt in self._enter:
compute_timestamps(edge.tgt, timestamp)
if self._undirected:
for edge in self._graph.find_node(node).incoming_edges:
if not edge.src in self._enter:
compute_timestamps(edge.src, timestamp)
self._leave[node] = next(timestamp)
timestamp = itertools.count()
for node in self._graph.nodes:
if not node.id in self._enter:
compute_timestamps(node.id, timestamp)
self.n_runs += 1
def is_back_edge(self, edge):
return \
self._enter[edge.tgt] < self._enter[edge.src] and \
self._leave[edge.src] < self._leave[edge.tgt]
class InspectedGraph(object):
def __init__(self, graph):
self.graph = graph
self.n_nodes = len(graph.nodes)
self.dfs = DepthFirstSearch(graph)
self.undirected_dfs = DepthFirstSearch(graph, undirected=True)
def n_root_nodes(self):
return sum(1 for node in self.graph.nodes if node.is_root())
def n_leaf_nodes(self):
return sum(1 for node in self.graph.nodes if node.is_leaf())
def n_top_nodes(self):
return sum(1 for node in self.graph.nodes if node.is_top())
def n_singleton_nodes(self):
return sum(1 for node in self.graph.nodes if node.is_singleton())
def n_loops(self):
return sum(1 for edge in self.graph.edges if edge.is_loop())
def n_components(self):
return self.undirected_dfs.n_runs - self.n_singleton_nodes()
def is_cyclic(self):
for edge in self.graph.edges:
if edge.is_loop() or self.dfs.is_back_edge(edge):
return True
return False
def is_forest(self):
if self.is_cyclic():
return False
else:
for node in self.graph.nodes:
if len(node.incoming_edges) > 1:
return False
return True
def is_tree(self):
return self.is_forest() and self.n_components() == 1
def treewidth(self):
n_nodes = len(self.graph.nodes) - self.n_singleton_nodes()
if n_nodes <= 1:
return 1
else:
undirected_graph = {}
for node in self.graph.nodes:
if not node.is_singleton():
undirected_graph[node.id] = set()
for edge in self.graph.edges:
if not edge.is_loop():
undirected_graph[edge.src].add(edge.tgt)
undirected_graph[edge.tgt].add(edge.src)
decomposition = quickbb(undirected_graph)
return max(1, max(len(u)-1 for u in decomposition))
def _crossing_pairs(self):
def endpoints(edge):
return (min(edge.src, edge.tgt), max(edge.src, edge.tgt))
for edge1 in self.graph.edges:
min1, max1 = endpoints(edge1)
for edge2 in self.graph.edges:
min2, max2 = endpoints(edge2)
if min1 < min2 and min2 < max1 and max1 < max2:
yield (min1, max1), (min2, max2)
def _crossing_edges(self):
crossing_edges = set()
for edge1, edge2 in self._crossing_pairs():
crossing_edges.add(edge1)
crossing_edges.add(edge2)
return crossing_edges
def is_noncrossing(self):
for _, _ in self._crossing_pairs():
return False
return True
def is_page2(self):
crossing_graph = {u: set() for u in self._crossing_edges()}
for edge1, edge2 in self._crossing_pairs():
crossing_graph[edge1].add(edge2)
crossing_graph[edge2].add(edge1)
# Tests whether the specified undirected graph is 2-colorable.
colors = {}
def inner(node, color1, color2):
colors[node] = color1
for neighbour in crossing_graph[node]:
if neighbour in colors:
if colors[neighbour] == color1:
return False
else:
inner(neighbour, color2, color1)
return True
for node in crossing_graph:
if node not in colors:
if not inner(node, 0, 1):
return False
return True
def density(self):
n_nodes = len(self.graph.nodes) - self.n_singleton_nodes()
if n_nodes <= 1:
return 1
else:
n_edges = 0
for edge in self.graph.edges:
if edge.src != edge.tgt:
n_edges += 1
return n_edges / (n_nodes - 1)
PROPERTY_COUNTER = itertools.count(1)
def report(msg, val):
print("(%02d)\t%s\t%s" % (next(PROPERTY_COUNTER), msg, val))
def analyze(graphs, ids=None):
ordered = False
n_graphs = 0
n_graphs_noncrossing = 0
n_graphs_has_top_node = 0
n_graphs_multirooted = 0
n_nodes = 0
n_nodes_with_reentrancies = 0
n_singletons = 0
n_top_nodes = 0
n_edges = 0
n_labels = 0;
n_properties = 0;
n_anchors = 0;
n_attributes = 0;
n_loops = 0
labels = set()
non_functional_labels = set()
n_cyclic = 0
n_connected = 0
n_forests = 0
n_trees = 0
n_graphs_page2 = 0
acc_treewidth = 0
n_roots_nontop = 0
acc_density = 0.0
max_treewidth = 0
acc_edge_length = 0
n_treewidth_one = 0
treewidths = []
for graph in graphs:
if ids and not graph.id in ids:
continue
n_graphs += 1
n_nodes += len(graph.nodes)
n_edges += len(graph.edges)
for node in graph.nodes:
if node.label is not None: n_labels += 1;
if node.properties is not None and node.values is not None:
n_properties += len(node.properties);
if node.anchors is not None: n_anchors += 1;
for edge in graph.edges:
if edge.attributes is not None and edge.values is not None:
n_attributes += len(edge.attributes);
inspected_graph = InspectedGraph(graph)
treewidth = inspected_graph.treewidth()
n_trees += inspected_graph.is_tree()
acc_density += inspected_graph.density()
has_reentrancies = False
has_top_node = False
n_loops += inspected_graph.n_loops()
for edge in graph.edges:
if edge.lab is not None: labels.add(edge.lab)
for node in graph.nodes:
n_top_nodes += node.is_top
if node.is_top:
has_top_node = True
n_singletons += node.is_singleton()
if len(node.incoming_edges) > 1:
n_nodes_with_reentrancies += 1
has_reentrancies = True
outgoing_labels = set()
for edge in node.outgoing_edges:
if edge.lab in outgoing_labels:
non_functional_labels.add(edge.lab)
else:
outgoing_labels.add(edge.lab)
if not node.is_singleton() and node.is_root() and not node.is_top:
n_roots_nontop += 1
n_cyclic += inspected_graph.is_cyclic()
n_connected += inspected_graph.n_components() == 1
n_forests += inspected_graph.is_forest()
acc_treewidth += treewidth
max_treewidth = max(max_treewidth, treewidth)
n_treewidth_one += treewidth == 1
treewidths.append(treewidth)
if graph.flavor == 0:
ordered = True
n_graphs_noncrossing += inspected_graph.is_noncrossing()
n_graphs_page2 += inspected_graph.is_page2()
acc_edge_length += sum(edge.length() for edge in graph.edges)
else:
if ordered:
print(
"analyzer.py: cannot mix graphs of different flavors in one file; exit.", file=sys.stderr)
sys.exit(1)
n_graphs_has_top_node += has_top_node
n_graphs_multirooted += inspected_graph.n_root_nodes() > 1
n_nonsingletons = n_nodes - n_singletons
report("number of graphs", "%d" % n_graphs)
report("number of nodes", "%d" % n_nodes)
n_tuples = n_top_nodes + n_labels + n_properties + n_anchors + n_edges + n_attributes;
if n_tuples > 0:
report("number of tops (percentage)",
"{:d} ({:.2f})".format(n_top_nodes, 100 * n_top_nodes / n_tuples));
report("number of node labels (percentage)",
"{:d} ({:.2f})".format(n_labels, 100 * n_labels / n_tuples));
report("number of node properties (percentage)",
"{:d} ({:.2f})".format(n_properties, 100 * n_properties / n_tuples));
report("number of node anchors (percentage)",
"{:d} ({:.2f})".format(n_anchors, 100 * n_anchors / n_tuples));
report("number of edges (percentage)",
"{:d} ({:.2f})".format(n_edges, 100 * n_edges / n_tuples));
report("number of edge attributes (percentage)",
"{:d} ({:.2f})".format(n_attributes, 100 * n_attributes / n_tuples));
report("number of edge labels", "%d" % len(labels))
# report("\\percentnode\\ singleton", "%.2f" % (100 * n_singletons / n_nodes))
# report("\\percentnode\\ non-singleton", "%.2f" % (100 * n_nonsingletons / n_nodes))
report("\\percentgraph\\ trees", "%.2f" % (100 * n_trees / n_graphs))
report("\\percentgraph\\ treewidth one", "%.2f" %
(100 * n_treewidth_one / n_graphs))
report("average treewidth", "%.3f" % (acc_treewidth / n_graphs))
# report("median treewidth", "%d" % statistics.median(treewidths))
report("maximal treewidth", "%d" % max_treewidth)
# report("edge density", "%.3f" % (n_edges / n_nonsingletons))
report("average edge density", "%.3f" % (acc_density / n_graphs))
report("\\percentnode\\ reentrant", "%.2f" %
(100 * n_nodes_with_reentrancies / n_nonsingletons))
# report("labels", " ".join(sorted(labels)))
# report("functional labels", " ".join(sorted(labels - non_functional_labels)))
# report("non-functional labels", " ".join(sorted(non_functional_labels)))
# report("\\percentgraph\\ forests", "%.2f" % (100 * n_forests / n_graphs))
# report("number of top nodes", "%d" % n_top_nodes)
report("\\percentgraph\\ cyclic", "%.2f" % (100 * n_cyclic / n_graphs))
# report("number of self-loops", "%d" % n_loops)
report("\\percentgraph\\ not connected", "%.2f" %
(100 * (n_graphs - n_connected) / n_graphs))
# report("\\percentgraph\\ without top", "%.2f" % (100 * (n_graphs - n_graphs_has_top_node) / n_graphs))
# report("average top nodes per graph", "%.3f" % (n_top_nodes / n_graphs))
report("\\percentgraph\\ multi-rooted", "%.2f" %
(100 * n_graphs_multirooted / n_graphs))
report("percentage of non-top roots", "%.2f" %
(100 * n_roots_nontop / n_nonsingletons))
if ordered:
report("average edge length", "%.3f" % (acc_edge_length / n_edges))
report("\\percentgraph\\ noncrossing", "%.2f" %
(100 * n_graphs_noncrossing / n_graphs))
report("\\percentgraph\\ pagenumber two", "%.2f" %
(100 * n_graphs_page2 / n_graphs))
else:
report("average edge length", "--")
report("\\percentgraph\\ noncrossing", "--")
report("\\percentgraph\\ pagenumber two", "--")
def read_ids(file_name):
ids = set()
with open(file_name) as fp:
for line in fp:
ids.add(line.rstrip())
return ids
def read_tokens(file_name):
with open(file_name) as fp:
for line in fp:
yield line.split()
def analyze_cmd(read_function, ordered=False):
import sys
ids = None
tokens = None
for arg in sys.argv[2:]:
x, y = tuple(arg.split(':'))
if x == 'ids':
print("Reading whitelisted IDs from %s" % y, file=sys.stderr)
ids = read_ids(y)
if x == 'tokens':
print("Reading tokens from %s" % y, file=sys.stderr)
tokens = read_tokens(y)
with open(sys.argv[1]) as fp:
analyze(read_function(fp), ordered=ordered, ids=ids, tokens=tokens)
|