Merge remote-tracking branch 'origin' into forceatlas2-networkx

.github/workflows/test.yml

@@ -143,7 +143,11 @@ jobs:
         run: |
           pip install --upgrade pip
           pip install -r requirements/default.txt -r requirements/test.txt
-          pip install --global-option=build_ext --global-option="-I/usr/local/include/" --global-option="-L/usr/local/lib/" pygraphviz
+          pip install --no-cache-dir \
+                      --config-settings="--global-option=build_ext" \
+                      --config-settings="--global-option=-I$(brew --prefix graphviz)/include/" \
+                      --config-settings="--global-option=-L$(brew --prefix graphviz)/lib/" \
+                      pygraphviz
           pip install -r requirements/extra.txt
           pip install .
           pip list

benchmarks/benchmarks/benchmark_to_networkx_graph.py

@@ -0,0 +1,30 @@
+import networkx as nx
+
+
+class ToNetworkXGraphBenchmark:
+    params = [nx.Graph, nx.DiGraph]
+    param_names = ["graph_type"]
+
+    def setup(self, graph_type):
+        self.edges = [(0, 1), (1, 2), (2, 3), (3, 4), (4, 5), (5, 6)]
+
+    def time_to_networkx_graph_direct(self, graph_type):
+        _ = nx.to_networkx_graph(self.edges, create_using=graph_type)
+
+    def time_to_networkx_graph_via_constructor(self, graph_type):
+        _ = graph_type(self.edges)
+
+    ### NOTE: Multi-instance checks are explicitly included to cover the case
+    # where many graph instances are created, which is not uncommon in graph
+    # analysis. The reason why multi-instance is explicitly probed (rather than
+    # relying solely on the number of repeats/runs from `timeit` in the benchmark
+    # suite) is to capture/amplify any distinctions from potential import
+    # cacheing of the try-excepts in the *same* run
+
+    def time_to_networkx_graph_direct_multi_instance(self, graph_type):
+        for _ in range(500):  # Creating many graph instances
+            _ = nx.to_networkx_graph(self.edges, create_using=graph_type)
+
+    def time_to_networkx_graph_via_constructor_multi_instance(self, graph_type):
+        for _ in range(500):  # Creating many graph instances
+            _ = graph_type(self.edges)

doc/developer/about_us.rst

@@ -153,6 +153,7 @@ to add your name to the bottom of the list.
 - Sebastiano Vigna, Github: `https://github.com/vigna`
 - Aaron Zolnai-Lucas, GitHub: `aaronzo <https://github.com/aaronzo>`_, LinkedIn: `aaronzolnailucas <https://www.linkedin.com/in/aaronzolnailucas/>`_
 - Erik Welch, GitHub: `eriknw <https://github.com/eriknw>`_, LinkedIn: `eriknwelch <https://www.linkedin.com/in/eriknwelch/>`_
+- Mohamed Rezk, Github: `mohamedrezk122 <https://github.com/mohamedrezk122>`_
 
 A supplementary (but still incomplete) list of contributors is given by the
 list of names that have commits in ``networkx``'s

examples/algorithms/plot_lca.py

@@ -14,34 +14,44 @@
 import networkx as nx
 import matplotlib.pyplot as plt
 
-# Generate a random tree with its node positions
-G = nx.random_tree(14, seed=100, create_using=nx.DiGraph)
+G = nx.DiGraph(
+    [
+        (0, 2),
+        (2, 7),
+        (2, 11),
+        (5, 6),
+        (6, 9),
+        (6, 8),
+        (7, 1),
+        (7, 3),
+        (8, 12),
+        (11, 10),
+        (11, 5),
+        (12, 4),
+        (12, 13),
+    ]
+)
 pos = nx.nx_agraph.graphviz_layout(G, prog="dot")
 
 # Compute lowest-common ancestors for certain node pairs
 ancestors = list(nx.all_pairs_lowest_common_ancestor(G, ((1, 3), (4, 9), (13, 10))))
 
 # Create node color and edge color lists
-node_color_map = []
-edge_color_map = []
-for i in range(14):
-    node_color_map.append("#D5D7D8")
-    edge_color_map.append("None")
-template = ["#FFE799", "#FFD23F", "#CEB6E2", "#A77CCB", "#88DFE7", "#45CDD9"]
-x = 0
-for i in ancestors:
-    for j in i[0]:
-        node_color_map[j] = template[x]
-    x += 1
-    node_color_map[i[1]] = template[x]
-    edge_color_map[i[1]] = "black"
-    x += 1
+node_colors = ["#D5D7D8" for _ in G]
+node_edge_colors = ["None" for _ in G]
+node_seq = list(G.nodes)
+clr_pairs = (("cyan", "tab:blue"), ("moccasin", "tab:orange"), ("lime", "tab:green"))
+for (children, ancestor), (child_clr, anc_clr) in zip(ancestors, clr_pairs):
+    for c in children:
+        node_colors[node_seq.index(c)] = child_clr
+    node_colors[node_seq.index(ancestor)] = anc_clr
+    node_edge_colors[node_seq.index(ancestor)] = "black"
 
 # Plot tree
 plt.figure(figsize=(15, 15))
 plt.title("Visualize Lowest Common Ancestors of node pairs")
 nx.draw_networkx_nodes(
-    G, pos, node_color=node_color_map, node_size=2000, edgecolors=edge_color_map
+    G, pos, node_color=node_colors, node_size=2000, edgecolors=node_edge_colors
 )
 nx.draw_networkx_edges(G, pos)
 nx.draw_networkx_labels(G, pos, font_size=15)

networkx/algorithms/non_randomness.py

@@ -51,7 +51,7 @@ def non_randomness(G, k=None, weight="weight"):
     NetworkXException
         if the input graph is not connected.
     NetworkXError
-        if the input graph contains self-loops.
+        if the input graph contains self-loops or if graph has no edges.
 
     Examples
     --------
@@ -74,6 +74,9 @@ def non_randomness(G, k=None, weight="weight"):
     """
     import numpy as np
 
+    # corner case: graph has no edges
+    if nx.is_empty(G):
+        raise nx.NetworkXError("non_randomness not applicable to empty graphs")
     if not nx.is_connected(G):
         raise nx.NetworkXException("Non connected graph.")
     if len(list(nx.selfloop_edges(G))) > 0:

networkx/algorithms/tests/test_cycles.py

@@ -947,7 +947,7 @@ class TestGirth:
             (nx.petersen_graph(), 5),
             (nx.heawood_graph(), 6),
             (nx.pappus_graph(), 6),
-            (nx.random_tree(10, seed=42), inf),
+            (nx.random_labeled_tree(10, seed=42), inf),
             (nx.empty_graph(10), inf),
             (nx.Graph(chain(cycle_edges(range(5)), cycle_edges(range(6, 10)))), 4),
             (

networkx/algorithms/tests/test_non_randomness.py

@@ -22,16 +22,21 @@ def test_non_randomness(k, weight, expected):
 
 
 def test_non_connected():
-    G = nx.Graph()
-    G.add_edge(1, 2)
+    G = nx.Graph([(1, 2)])
     G.add_node(3)
-    with pytest.raises(nx.NetworkXException):
+    with pytest.raises(nx.NetworkXException, match="Non connected"):
         nx.non_randomness(G)
 
 
 def test_self_loops():
     G = nx.Graph()
     G.add_edge(1, 2)
     G.add_edge(1, 1)
-    with pytest.raises(nx.NetworkXError):
+    with pytest.raises(nx.NetworkXError, match="Graph must not contain self-loops"):
+        nx.non_randomness(G)
+
+
+def test_empty_graph():
+    G = nx.empty_graph(1)
+    with pytest.raises(nx.NetworkXError, match=".*not applicable to empty graphs"):
         nx.non_randomness(G)

networkx/algorithms/tests/test_swap.py

@@ -3,7 +3,8 @@
 import networkx as nx
 
 cycle = nx.cycle_graph(5, create_using=nx.DiGraph)
-tree = nx.random_tree(10, create_using=nx.DiGraph, seed=42)
+tree = nx.DiGraph()
+tree.add_edges_from(nx.random_labeled_tree(10, seed=42).edges)
 path = nx.path_graph(5, create_using=nx.DiGraph)
 binomial = nx.binomial_tree(3, create_using=nx.DiGraph)
 HH = nx.directed_havel_hakimi_graph([1, 2, 1, 2, 2, 2], [3, 1, 0, 1, 2, 3])

networkx/classes/reportviews.py

@@ -82,6 +82,7 @@
 
     The argument `nbunch` restricts edges to those incident to nodes in nbunch.
 """
+from abc import ABC
 from collections.abc import Mapping, Set
 
 import networkx as nx
@@ -734,8 +735,15 @@ def __iter__(self):
                 yield (n, deg)
 
 
+# A base class for all edge views. Ensures all edge view and edge data view
+# objects/classes are captured by `isinstance(obj, EdgeViewABC)` and
+# `issubclass(cls, EdgeViewABC)` respectively
+class EdgeViewABC(ABC):
+    pass
+
+
 # EdgeDataViews
-class OutEdgeDataView:
+class OutEdgeDataView(EdgeViewABC):
     """EdgeDataView for outward edges of DiGraph; See EdgeDataView"""
 
     __slots__ = (
@@ -1036,7 +1044,7 @@ def __contains__(self, e):
 
 
 # EdgeViews    have set operations and no data reported
-class OutEdgeView(Set, Mapping):
+class OutEdgeView(Set, Mapping, EdgeViewABC):
     """A EdgeView class for outward edges of a DiGraph"""
 
     __slots__ = ("_adjdict", "_graph", "_nodes_nbrs")

networkx/classes/tests/test_reportviews.py

@@ -1425,3 +1425,11 @@ def test_cache_dict_get_set_state(graph):
     # Raises error if the cached properties and views do not work
     pickle.loads(pickle.dumps(G, -1))
     deepcopy(G)
+
+
+def test_edge_views_inherit_from_EdgeViewABC():
+    all_edge_view_classes = (v for v in dir(nx.reportviews) if "Edge" in v)
+    for eview_class in all_edge_view_classes:
+        assert issubclass(
+            getattr(nx.reportviews, eview_class), nx.reportviews.EdgeViewABC
+        )

networkx/convert.py

@@ -90,13 +90,6 @@ def to_networkx_graph(data, create_using=None, multigraph_input=False):
         except Exception as err:
             raise nx.NetworkXError("Input is not a correct NetworkX graph.") from err
 
-    # pygraphviz  agraph
-    if hasattr(data, "is_strict"):
-        try:
-            return nx.nx_agraph.from_agraph(data, create_using=create_using)
-        except Exception as err:
-            raise nx.NetworkXError("Input is not a correct pygraphviz graph.") from err
-
     # dict of dicts/lists
     if isinstance(data, dict):
         try:
@@ -113,6 +106,20 @@ def to_networkx_graph(data, create_using=None, multigraph_input=False):
             except Exception as err2:
                 raise TypeError("Input is not known type.") from err2
 
+    # edgelists
+    if isinstance(data, list | tuple | nx.reportviews.EdgeViewABC | Iterator):
+        try:
+            return from_edgelist(data, create_using=create_using)
+        except:
+            pass
+
+    # pygraphviz  agraph
+    if hasattr(data, "is_strict"):
+        try:
+            return nx.nx_agraph.from_agraph(data, create_using=create_using)
+        except Exception as err:
+            raise nx.NetworkXError("Input is not a correct pygraphviz graph.") from err
+
     # Pandas DataFrame
     try:
         import pandas as pd

networkx/tests/test_all_random_functions.py

@@ -194,7 +194,7 @@ def kernel_integral(u, w, z):
     t(nx.random_lobster, n, p1, p2, seed=seed)
     t(nx.random_powerlaw_tree, n, seed=seed, tries=5000)
     t(nx.random_powerlaw_tree_sequence, 10, seed=seed, tries=5000)
-    t(nx.random_tree, n, seed=seed)
+    t(nx.random_labeled_tree, n, seed=seed)
     t(nx.utils.powerlaw_sequence, n, seed=seed)
     t(nx.utils.zipf_rv, 2.3, seed=seed)
     cdist = [0.2, 0.4, 0.5, 0.7, 0.9, 1.0]

pyproject.toml

@@ -67,13 +67,13 @@ developer = [
     'rtoml',
 ]
 doc = [
-    'sphinx>=7',
-    'pydata-sphinx-theme>=0.14',
-    'sphinx-gallery>=0.14',
+    'sphinx>=7.3',
+    'pydata-sphinx-theme>=0.15',
+    'sphinx-gallery>=0.16',
     'numpydoc>=1.7',
     'pillow>=9.4',
     'texext>=0.6.7',
-    'myst-nb>=1.0',
+    'myst-nb>=1.1',
 ]
 extra = [
     'lxml>=4.6',

requirements/doc.txt

@@ -1,9 +1,9 @@
 # Generated via tools/generate_requirements.py and pre-commit hook.
 # Do not edit this file; modify pyproject.toml instead.
-sphinx>=7
-pydata-sphinx-theme>=0.14
-sphinx-gallery>=0.14
+sphinx>=7.3
+pydata-sphinx-theme>=0.15
+sphinx-gallery>=0.16
 numpydoc>=1.7
 pillow>=9.4
 texext>=0.6.7
-myst-nb>=1.0
+myst-nb>=1.1