first commit, isolates and betweenness

.github/workflows/test.yml

@@ -32,9 +32,10 @@ jobs:
         run: |
           conda install -c conda-forge joblib scipy pandas pytest-cov pytest-randomly
             # matplotlib lxml pygraphviz pydot sympy  # Extra networkx deps we don't need yet
-          pip install git+https://github.com/networkx/networkx.git@main --no-deps
-          pip install -e . --no-deps
+          python -m pip install git+https://github.com/networkx/networkx.git@main
+          python -m pip install .
           echo "Done with installing"
       - name: PyTest
         run: |
           NETWORKX_GRAPH_CONVERT=parallel pytest --pyargs networkx
+          python -m pytest --pyargs nx_parallel

.gitignore

@@ -127,4 +127,3 @@ dmypy.json
 
 # Pyre type checker
 .pyre/
-

.pre-commit-config.yaml

@@ -0,0 +1,31 @@
+# Install pre-commit hooks via
+# pre-commit install
+
+repos:
+  - repo: https://github.com/psf/black
+    rev: 23.3.0
+    hooks:
+      - id: black
+  - repo: https://github.com/adamchainz/blacken-docs
+    rev: 1.13.0
+    hooks:
+      - id: blacken-docs
+  - repo: https://github.com/pre-commit/mirrors-prettier
+    rev: v2.7.1
+    hooks:
+      - id: prettier
+        files: \.(html|md|toml|yml|yaml)
+        args: [--prose-wrap=preserve]
+  - repo: https://github.com/charliermarsh/ruff-pre-commit
+    rev: v0.0.258
+    hooks:
+      - id: ruff
+        args:
+          - --fix
+  - repo: local
+    hooks:
+      - id: pyproject.toml
+        name: pyproject.toml
+        language: system
+        entry: python tools/generate_pyproject.toml.py
+        files: "pyproject.toml|requirements/.*\\.txt|tools/.*pyproject.*"

README.md

@@ -1,7 +1,7 @@
-NX-Parallel
+nx_parallel
 -----------
 
-A NetworkX backend plugin which uses dask for parallelization.
+A NetworkX backend plugin which uses joblib and multiprocessing for parallelization.
 
 ``` python
 In [1]: import networkx as nx; import nx_parallel
@@ -23,4 +23,19 @@ Out[4]:
  8: 0.0,
  9: 0.0}
 
-```
+```
+
+Currently the following functions have parallelized implementations:
+  - centrality
+    - betweenness_centrality
+  - tournament
+    - is_reachable
+  - closeness_vitality
+  - efficiency_measures
+    - local_efficiency
+
+![alt text](timing/heatmap_all_functions.png)
+
+See the ```/timing``` folder for more heatmaps and code for heatmap generation!
+
+

nx_parallel/__init__.py

@@ -1,3 +1,3 @@
-from .centrality import *
-from .graph import *
+from .algorithms import *
+from .classes import *
 from .interface import *

nx_parallel/algorithms/__init__.py

@@ -0,0 +1,8 @@
+# subpackages
+from .centrality import *
+from .utils import *
+
+# modules
+from .efficiency_measures import *
+from .isolate import *
+from .tournament import *

nx_parallel/algorithms/centrality/__init__.py

@@ -0,0 +1 @@
+from .betweenness import *

nx_parallel/algorithms/centrality/betweenness.py

@@ -0,0 +1,121 @@
+from joblib import Parallel, delayed, cpu_count
+from nx_parallel.algorithms.utils.chunk import chunks
+from networkx.utils import py_random_state
+from networkx.algorithms.centrality.betweenness import (
+    _rescale,
+    _single_source_shortest_path_basic,
+    _single_source_dijkstra_path_basic,
+    _accumulate_endpoints,
+    _accumulate_basic,
+)
+
+__all__ = ["betweenness_centrality"]
+
+
+@py_random_state(5)
+def betweenness_centrality(
+    G, k=None, normalized=True, weight=None, endpoints=False, seed=None
+):
+    r"""Parallel Compute shortest-path betweenness centrality for nodes
+
+    Betweenness centrality of a node $v$ is the sum of the
+    fraction of all-pairs shortest paths that pass through $v$
+
+    .. math::
+
+       c_B(v) =\sum_{s,t \in V} \frac{\sigma(s, t|v)}{\sigma(s, t)}
+
+    where $V$ is the set of nodes, $\sigma(s, t)$ is the number of
+    shortest $(s, t)$-paths,  and $\sigma(s, t|v)$ is the number of
+    those paths  passing through some  node $v$ other than $s, t$.
+    If $s = t$, $\sigma(s, t) = 1$, and if $v \in {s, t}$,
+    $\sigma(s, t|v) = 0$ [2]_.
+
+    Parameters
+    ----------
+    G : graph
+      A NetworkX graph.
+
+    k : int, optional (default=None)
+      If k is not None use k node samples to estimate betweenness.
+      The value of k <= n where n is the number of nodes in the graph.
+      Higher values give better approximation.
+
+    normalized : bool, optional
+      If True the betweenness values are normalized by `2/((n-1)(n-2))`
+      for graphs, and `1/((n-1)(n-2))` for directed graphs where `n`
+      is the number of nodes in G.
+
+    weight : None or string, optional (default=None)
+      If None, all edge weights are considered equal.
+      Otherwise holds the name of the edge attribute used as weight.
+      Weights are used to calculate weighted shortest paths, so they are
+      interpreted as distances.
+
+    endpoints : bool, optional
+      If True include the endpoints in the shortest path counts.
+
+    seed : integer, random_state, or None (default)
+        Indicator of random number generation state.
+        See :ref:`Randomness<randomness>`.
+        Note that this is only used if k is not None.
+
+    Returns
+    -------
+    nodes : dictionary
+       Dictionary of nodes with betweenness centrality as the value.
+
+    Notes
+    -----
+    This algorithm is a parallelized version of betwenness centrality in NetworkX.
+    Nodes are divided into chunks based on the number of available processors,
+    and otherwise all calculations are similar.
+    """
+    if k is None:
+        nodes = G.nodes
+    else:
+        nodes = seed.sample(list(G.nodes), k)
+    total_cores = cpu_count()
+    num_chunks = max(len(nodes) // total_cores, 1)
+    node_chunks = list(chunks(nodes, num_chunks))
+    bt_cs = Parallel(n_jobs=total_cores)(
+        delayed(betweenness_centrality_node_subset)(
+            G,
+            chunk,
+            weight,
+            endpoints,
+        )
+        for chunk in node_chunks
+    )
+
+    # Reducing partial solution
+    bt_c = bt_cs[0]
+    for bt in bt_cs[1:]:
+        for n in bt:
+            bt_c[n] += bt[n]
+
+    betweenness = _rescale(
+        bt_c,
+        len(G),
+        normalized=normalized,
+        directed=G.is_directed(),
+        k=k,
+        endpoints=endpoints,
+    )
+    return betweenness
+
+
+def betweenness_centrality_node_subset(G, nodes, weight=None, endpoints=False):
+    betweenness = dict.fromkeys(G, 0.0)
+    for s in nodes:
+        # single source shortest paths
+        if weight is None:  # use BFS
+            S, P, sigma, _ = _single_source_shortest_path_basic(G, s)
+        else:  # use Dijkstra's algorithm
+            S, P, sigma, _ = _single_source_dijkstra_path_basic(G, s, weight)
+        # accumulation
+        if endpoints:
+            betweenness, delta = _accumulate_endpoints(betweenness, S, P, sigma, s)
+        else:
+            betweenness, delta = _accumulate_basic(betweenness, S, P, sigma, s)
+    return betweenness