Merge pull request #138 from benedekrozemberczki/grarep_warning

Now we clone the graph before adding the main diagonal, with a proper warning

karateclub/estimator.py

@@ -1,7 +1,9 @@
 import random
 import numpy as np
 import networkx as nx
+import warnings
 from typing import List
+from tqdm.auto import trange
 import re
 
 """General Estimator base class."""
@@ -45,10 +47,48 @@ def _set_seed(self):
  np.random.seed(self.seed)
 
  @staticmethod
- def _ensure_integrity(graph: nx.classes.graph.Graph) -> nx.classes.graph.Graph:
+ def _ensure_walk_traversal_conditions(graph: nx.classes.graph.Graph) -> nx.classes.graph.Graph:
  """Ensure walk traversal conditions."""
- edge_list = [(index, index) for index in range(graph.number_of_nodes())]
- graph.add_edges_from(edge_list)
+ for node_index in trange(
+ graph.number_of_nodes(),
+ # We do not leave the bar.
+ leave=False,
+ # We only show this bar when we can expect
+ # for this process to take a bit of time.
+ disable=graph.number_of_nodes() < 10_000,
+ desc="Checking main diagonal existance",
+ dynamic_ncols=True
+ ):
+ if not graph.has_edge(node_index, node_index):
+ warnings.warn(
+ (
+ "Please do be advised that "
+ "the graph you have provided does not "
+ "contain (some) edges in the main "
+ "diagonal, for instance the self-loop "
+ "constitued of ({}, {}). These selfloops "
+ "are necessary to ensure that the graph "
+ "is traversable, and for this reason we "
+ "create a copy of the graph and add therein "
+ "the missing edges. Since we are creating "
+ "a copy, this will immediately duplicate "
+ "the memory requirements. To avoid this double "
+ "allocation, you can provide the graph with the selfloops."
+ ).format(
+ node_index,
+ node_index
+ )
+ )
+ # We create a copy of the graph
+ graph = graph.copy()
+ # And we add the missing edges
+ # for filling the main diagonal
+ graph.add_edges_from((
+ (index, index)
+ for index in range(graph.number_of_nodes())
+ if not graph.has_edge(index, index)
+ ))
+ break
 
  return graph
 
@@ -63,7 +103,7 @@ def _check_indexing(graph: nx.classes.graph.Graph):
  def _check_graph(self, graph: nx.classes.graph.Graph) -> nx.classes.graph.Graph:
  """Check the Karate Club assumptions about the graph."""
  self._check_indexing(graph)
- graph = self._ensure_integrity(graph)
+ graph = self._ensure_walk_traversal_conditions(graph)
 
  return graph
 

karateclub/node_embedding/neighbourhood/first_order_line.py

@@ -68,7 +68,7 @@ def fit(self, graph: nx.classes.graph.Graph):
 
  self.embedding = np.random.uniform(size=(number_of_nodes, self.dimensions))
 
- for epoch in trange(self.epochs, desc="Epochs", disable=not self.verbose):
+ for epoch in trange(self.epochs, desc="Epochs", disable=not self.verbose, leave=False):
  for _ in range(number_of_edges // self.mini_batch_size):
  gradient = np.zeros_like(self.embedding)
 

karateclub/node_embedding/neighbourhood/second_order_line.py

@@ -72,7 +72,7 @@ def fit(self, graph: nx.classes.graph.Graph):
  self.src_embedding = np.random.uniform(size=(number_of_nodes, self.dimensions // 2))
  self.dst_embedding = np.random.uniform(size=(number_of_nodes, self.dimensions // 2))
 
- for epoch in trange(self.epochs, desc="Epochs", disable=not self.verbose):
+ for epoch in trange(self.epochs, desc="Epochs", disable=not self.verbose, leave=False):
  for _ in range(number_of_edges // self.mini_batch_size):
  src_gradient = np.zeros_like(self.src_embedding)
  dst_gradient = np.zeros_like(self.dst_embedding)