Note

This is the documentation for the current state of the development branch of rustworkx. The documentation or APIs here can change prior to being released.

Release Notes#

0.15.1-64#

New Features#

The following methods now support sequences and generators as inputs, in addition to the existing support for lists: * add_nodes_from() and add_nodes_from() * add_edges_from() and add_edges_from() * add_edges_from_no_data() and add_edges_from_no_data() * extend_from_edge_list() and extend_from_edge_list() * extend_from_weighted_edge_list() and extend_from_weighted_edge_list()

Added a new function, johnson_simple_cycles, to the rustworkx-core crate. This function implements Johnson’s algorithm for finding all elementary cycles in a directed graph.

Added a new trait EdgeFindable to find an EdgeIndex from a graph given a pair of node indices.

Added a new trait EdgeRemovable to remove an edge from a graph by its EdgeIndex.

Added a new function, degree_centrality() which is used to compute the degree centrality for all nodes in a given graph. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.hexagonal_lattice_graph(4, 4)
centrality = rx.degree_centrality(graph)

# Generate a color list
colors = []
for node in graph.node_indices():
    centrality_score = centrality[node]
    graph[node] = centrality_score
    colors.append(centrality_score)
mpl_draw(
    graph,
    with_labels=True,
    node_color=colors,
    node_size=650,
    labels=lambda x: "{0:.2f}".format(x)
)

Added two new functions, in_degree_centrality() and out_degree_centrality() to calculate two special types of degree centrality for directed graphs.

Add rustworkx.immediate_dominators() function for computing immediate dominators of all nodes in a directed graph. This function mirrors the networkx.immediate_dominators function.

Add the rustworkx.dominance_frontiers() function to compute the dominance frontiers of all nodes in a directed graph. This function mirrors the networkx.dominance_frontiers function.

Added a new function, karate_club_graph() that returns Zachary’s Karate Club graph, commonly found in social network examples.

import rustworkx.generators
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.karate_club_graph()
layout = rustworkx.circular_layout(graph)
mpl_draw(graph, pos=layout)

Added a new method neighbors_undirected() to obtain the neighbors of a node in a directed graph, irrespective of the edge directionality.

Added the ability to read GraphML files that are compressed using gzip, with function read_graphml(). The extensions .graphmlz and .gz are automatically recognised, but the gzip decompression can be forced with the “compression” optional argument.

Upgrade Notes#

The minimum supported Python version for using rustworkx has been raised to Python 3.9. Python 3.8 has reached it’s end-of-life and will no longer be supported. To use rustworkx you will need to ensure you are using Python >=3.9.

Bug Fixes#

Fixed a bug introduced in version 0.15 where the edge colors specified as a list in calls to mpl_draw() were not having their order respected. Instead, the order of the colors was being shuffled. This has been restored and now the behavior should match that of 0.14.

Fixed a bug in the type hints for find_node_by_weight() and find_node_by_weight(). Refer to issue 1243 for more information.

Fixed a bug in the type hint for node_link_json(). Refer to issue 1243 for more information.

Fix typos detected by typos. Add spell checker invocations to the Nox lint session.

0.15.1#

Prelude#

This release is a bugfix patch release that fixes an inadvertent breaking API change for the graphviz_draw() function in the 0.15.0 release.

Bug Fixes#

Fixed an issue in the graphviz_draw(), PyGraph.to_dot(), and PyDiGraph.to_dot() which was incorrectly escaping strings when upgrading to 0.15.0. In earlier versions of rustworkx if you manually placed quotes in a string for an attr callback to get that to pass through to the output dot file, this was incorrectly being converted in rustworkx 0.15.0 to duplicate the quotes and escape them. For example, if you defined a callback like:
```
def color_node(_node):
    return {
        "color": '"#422952"'
    }
```
to set the color attribute in the output dot file with the string "#422952" (with the quotes), this was incorrectly being converted to "\"#422952\"". This no longer occurs, in rustworkx 0.16.0 there will likely be additional options exposed in graphviz_draw(), PyGraph.to_dot(), and PyDiGraph.to_dot() to expose further options around this.

Fixed two bugs in the node position calculation done by the generator functions hexagonal_lattice_graph() and directed_hexagonal_lattice_graph() when with_positions = True:
- Corrected a scale factor that made all the hexagons in the lattice irregular
- Corrected an indexing bug that positioned the nodes in the last column of the lattice incorrectly when periodic = False and cols is odd

0.15.0#

Prelude#

This is a new feature release of Rustworkx that adds many new features to the library. The highlights of this release are:

An expansion of functions in rustworkx-core that previously only existed in the Python API.

Expanded graph coloring algorithms

This release moves to using the Python Stable ABI, while this release officially supports Python 3.8 through 3.12, the published binaries should be compatible with future Python versions too. Although there is no guarantee provided about future versions. Additionally, the minimum supported Rust version for building rustworkx and more importantly rustworkx-core is now 1.70.0. Additionally, in this release the macOS arm64 platform has been promoted from Tier 4 to Tier 1.

New Features#

Added a method has_node() to the PyGraph and PyDiGraph classes to check if a node is in the graph.

Added a new function rustworkx_core::dag_algo::layers to rustworkx-core to get the layers of a directed acyclic graph. This is equivalent to the layers() function that existed in the Python API but now exposes it for Rust users too.

Added two new functions, from_node_link_json_file() and parse_node_link_json(), which are used to parse a node link json object and generate a rustworkx PyGraph or PyDiGraph object from it.

Added a new function ancestors() to the rustworkx_core::traversal module. That is a generic Rust implementation for the core rust library that provides the ancestors() function to Rust users.

Added a new function descendants() to the rustworkx_core::traversal module. That is a generic Rust implementation for the core rust library that provides the descendants() function to Rust users.

Added a new function bfs_predecessors() to the rustworkx_core::traversal module. That is a generic Rust implementation for the core rust library that provides the bfs_predecessors() function to Rust users.

Added a new function bfs_successors() to the rustworkx_core::traversal module. That is a generic Rust implementation for the core rust library that provides the bfs_successors() function to Rust users.

Added a new function collect_runs to rustworkx-core’s dag_algo module. Previously, the collect_runs() functionality for DAGs was only exposed via the Python interface. Now Rust users can take advantage of this functionality in rustworkx-core.

Added a function connected_subgraphs() to determine all connected subgraphs of size \(k\) in polynomial delay for undirected graphs. This improves upon the brute-force method by two orders of magnitude for sparse graphs such as heavy-hex, enabling addressing larger graphs and for a larger \(k\). The introduced method is based on “Enumerating Connected Induced Subgraphs: Improved Delay and Experimental Comparison” by Christian Komusiewicz and Frank Sommer. In particular, the procedure Simple is implemented. Possible runtime improvement can be gained by parallelization over each recursion or by following the discussion in Lemma 4 of above work and thus implementing intermediate sets \(X\) and \(P\) more efficiently.

Rustworkx functions that return custom iterable objects, such as PyDiGraph.node_indices(), now each have an associated custom iterator and reversed-iterator object for these. This provides a speedup of approximately 40% for iterating through the custom iterables.

These types are not directly nameable or constructable from Python space, and other than the performance improvement, the behavior should largely not be noticeable from Python space.

Added rustworkx.generators.dorogovtsev_goltsev_mendes_graph() that generates deterministic scale-free graphs using the Dorogovtsev-Goltsev-Mendes iterative procedure.
import rustworkx as rx from rustworkx.visualization import mpl_draw graph = rx.generators.dorogovtsev_goltsev_mendes_graph(2) mpl_draw(graph)

Added a new function, rustworkx_core::generators::dorogovtsev_goltsev_mendes_graph, to rustworkx-core that is used to generate deterministic scale-free graphs using the Dorogovtsev-Goltsev-Mendes iterative procedure.

Node contraction is now supported for petgraph types StableGraph and GraphMap in rustworkx-core. To use it, import one of the ContractNodes* traits from graph_ext and call the contract_nodes method on your graph.

All current petgraph data structures now support testing for parallel edges in rustworkx-core. To use this, import HasParallelEdgesDirected or HasParallelEdgesUndirected depending on your graph type, and call the has_parallel_edges method on your graph.

A new trait NodeRemovable has been added to graph_ext module in rustworkx-core which provides a consistent interface for performing node removal operations on petgraph types Graph, StableGraph, GraphMap, and MatrixGraph. To use it, import NodeRemovable from graph_ext.

Adds new random graph generator function, hyperbolic_random_graph() to sample the hyperbolic random graph model. For example:

import math

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.hyperbolic_random_graph(
  [[math.sinh(0.5), 0, 3.14159], [-math.sinh(1), 0, -3.14159]],
    2.55,
    None,
)
mpl_draw(graph)

Adds new function to the rustworkx-core module rustworkx_core::generators hyperbolic_random_graph() that samples the hyperbolic random graph model.

lexicographical_topological_sort() and TopologicalSorter now both accept an initial keyword argument, which can be used to limit the returned topological orderings to be only over the nodes that are dominated by the initial set. This can provide performance improvements by removing the need for a search over all graph nodes to determine the initial set of nodes with zero in degree; this is particularly relevant to TopologicalSorter, where the user may terminate the search after only examining part of the order.

Added a new function, is_semi_connected() which will check if a :class: ~rustworkx.PyDiGraph object is semi-connected.

Added a new function lexicographical_topological_sort to the rustworkx_core::dag_algo module. That is a generic Rust implementation for the core rust library that provides the lexicographical_topological_sort() function to Rust users.

Added a new function digraph_maximum_bisimulation() to compute the maximum bisimulation or relational coarsest partition of a graph. This function is based on the algorithm described in the publication “Three partition refinement algorithms” by Paige and Tarjan. This function receives a graph and returns a RelationalCoarsestPartition.

Added a new class RelationalCoarsestPartition to output the maximum bisimulation or relational coarsest partition of a graph. This class contains instances of IndexPartitionBlock and can be iterated over.

Added a new class IndexPartitionBlock to output a block of a node partition. This class is an iterator over node indices.

Added a new function collect_bicolor_runs to rustworkx-core’s dag_algo module. Previously, the collect_bicolor_runs() functionality for DAGs was only exposed via the Python interface. Now Rust users can take advantage of this functionality in rustworkx-core.

Added a new module dag_algo to rustworkx-core which contains a new function longest_path function to rustworkx-core. Previously the longest_path() functionality for DAGs was only exposed via the Python interface. Now Rust users can take advantage of this functionality in rustworkx-core.

Added two new keyword arguments, periodic an with_positions, to the generator functions hexagonal_lattice_graph() and directed_hexagonal_lattice_graph(). If periodic is set to True the boundaries of the lattice will be joined to form a periodic grid. If the with_positions argument is set to True than the data payload of all the nodes will be set to a tuple of the form (x, y) where x and y represent the node’s position in the lattice. For example:
import rustworkx as rx from rustworkx.visualization import mpl_draw graph = rx.generators.hexagonal_lattice_graph(4, 4, periodic=True, with_positions=True) mpl_draw(graph, with_labels=True, labels=str)

Added a new rustworkx-core function rustworkx_core::generators::hexagonal_lattice_graph_weighted() which is used to generate a hexagonal lattice graph where a callback is used to generate the node weights for each node from a tuple of the form (usize, usize).

Added the PyDiGraph.remove_node_retain_edges_by_id() and remove_node_retain_edges_by_key() methods, which provide a node-removal that is linear in the degree of the node, as opposed to quadratic like remove_node_retain_edges(). These methods require, respectively, that the edge weights are referentially identical if they should be retained (a is b, in Python), or that you can supply a key function that produces a Python-hashable result that is used to do the equality matching between input and output edges.

Added a new class ColoringStrategy used to specify the strategy used by the greedy node and edge coloring algorithms. The Degree strategy colors the nodes with higher degree first. The Saturation strategy dynamically chooses the vertex that has the largest number of different colors already assigned to its neighbors, and, in case of a tie, the vertex that has the largest number of uncolored neighbors. The IndependentSet strategy finds independent subsets of the graph, and assigns a different color to each of these subsets.

The rustworkx-core coloring module has 2 new functions, greedy_node_color_with_coloring_strategy and greedy_edge_color_with_coloring_strategy. These functions color respectively the nodes or the edges of the graph using the specified coloring strategy and handling the preset colors when provided.

Added a new keyword argument, strategy, to graph_greedy_color() and to graph_greedy_edge_color() to specify the greedy coloring strategy.

For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.generalized_petersen_graph(5, 2)

coloring = rx.graph_greedy_color(graph, strategy=rx.ColoringStrategy.Saturation)
colors = [coloring[node] for node in graph.node_indices()]

layout = rx.shell_layout(graph, nlist=[[0, 1, 2, 3, 4],[6, 7, 8, 9, 5]])
mpl_draw(graph, node_color=colors, pos=layout)

Added a new keyword argument, preset_color_fn, to graph_greedy_edge_color() which is used to provide preset colors for specific edges when computing the graph coloring. You can optionally pass a callable to that argument which will be passed edge index from the graph and is either expected to return an integer color to use for that edge, or None to indicate there is no preset color for that edge. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.generalized_petersen_graph(5, 2)

def preset_colors(edge_index):
    if edge_index == 0:
        return 3

coloring = rx.graph_greedy_edge_color(graph, preset_color_fn=preset_colors)
colors = [coloring[edge] for edge in graph.edge_indices()]

layout = rx.shell_layout(graph, nlist=[[0, 1, 2, 3, 4], [6, 7, 8, 9, 5]])
mpl_draw(graph, edge_color=colors, pos=layout)

Adds new random graph generator in rustworkx for the stochastic block model. There is a generator for directed directed_sbm_random_graph() and undirected graphs undirected_sbm_random_graph().

import numpy as np

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.undirected_sbm_random_graph(
    [2, 1],
    np.array([[1, 1], [1, 0]], dtype=float),
    False,
)
mpl_draw(graph)

Adds new function sbm_random_graph to the rustworkx-core module rustworkx_core::generators that samples a graph from the stochastic block model.

rustworkx wheels are now built against Python’s stable Application Binary Interface (ABI). For rustworkx users, this means that wheels distributed by us via PyPI will continue to work with newer versions of Python without having to recompile the code. This change will also simplify the release process for the developers and reduce the storage size required to mirror rustworkx wheels.

TopologicalSorter now has a check_args keyword argument, which can be set to False to disable the runtime detection of invalid arguments to done(). This provides a memory and runtime improvement to the online sorter, at the cost that the results will be undefined and likely meaningless if invalid values are given.

TopologicalSorter.done() now accepts single integers, in addition to lists of integers. This can be a sizeable performance improvement for algorithms that iterate through nodes, and only conditionally mark them as done; there is no longer a need to allocate a temporary Python array.

lexicographical_topological_sort() and TopologicalSorter now accept a reverse keyword argument, which can be set to True to find a “reversed” topological ordering. This is a topological ordering that would be found if all the edges in the graph had their directions reversed.

PyGraph and PyDiGraph now each take keyword arguments node_count_hint and edge_count_hint in their constructors, which can be used to pre-allocate space for the given number of nodes and edges.

Upgrade Notes#

The minimum supported rust version to build rustworkx and rustworkx-core has been raised from 1.64 to 1.70. You will need to upgrade your rust compiler version to at least version 1.70 to continue building from source. Python library users who are installing rustworkx on a supported platform will not need to make any changes.

The rustworkx-core function rustworkx_core::connectivity::find_cycle now requires the petgraph::visit::Visitable trait for its input argument graph. This was required to fix the behavior when source is None to ensure we always find a cycle if one exists.

The interface of the rustworkx_core::generators::hexagonal_lattice_graph() function has been changed, there is a new required boolean argument periodic which is used to indicate whether the output graph should join the bondaries of the lattice to form a periodic grid or not. This argument didn’t exist in prior releases of rustworkx-core and it will need to be added when upgrading to this new release.

Bug Fixes#

Fixed the behavior of digraph_find_cycle() when no source node was provided. Previously, the function would start looking for a cycle at an arbitrary node which was not guaranteed to return a cycle. Now, the function will smartly choose a source node to start the search from such that if a cycle exists, it will be found.

Fixed an issue with the graphviz_draw() where it would not correctly escape special characters in all scenarios. This has been corrected so you can now use special characters with the function, for example:

import rustworkx as rx
from rustworkx.visualization import graphviz_draw

graphviz_draw(
    rx.generators.path_graph(2),
    node_attr_fn=lambda x: {"label": "the\nlabel", "tooltip": "the\ntooltip"},
)

Fixed: #750

Fixed the plots of multigraphs using mpl_draw(). Previously, parallel edges of multigraphs were plotted on top of each other, with overlapping arrows and labels. The radius of parallel edges of the multigraph was fixed to be 0.25 for connectionstyle supporting this argument in draw_edges(). The edge labels were offset to 0.25 in draw_edge_labels() to align with their respective edges. This fix can be tested using the following code:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.PyDiGraph() graph.add_node('A') graph.add_node('B') graph.add_node('C') graph.add_edge(1, 0, 2) graph.add_edge(0, 1, 3) graph.add_edge(1, 2, 4) mpl_draw(graph, with_labels=True, labels=str, edge_labels=str, alpha=0.5)
Fixed #774

Fixed a bug in the type hint for the mpl_draw(). Previously, the type hint indicated that all kwargs were required when calling the method. The type annotation has been updated to indicate that kwargs with partial arguments is allowed.

Fixed an issue with the Dijkstra path functions:
where a PanicException was raised without much detail when an invalid node index was passed in to the source argument. This has been corrected so an IndexError is raised instead. Fixed #1117

Fixed the bug type hint for the bfs_search(), dfs_search() and dijkstra_search(). Refer to #1130 for more information.

Fixed support for handling Long type attributes from input GraphML in the read_graphml() function. Fixed #1140.

Other Notes#

Support for the arm64 macOS platform has been promoted from Tier 4 to Tier 1. Previously the platform was at Tier 4 because there was no available CI environment for testing rustworkx on the platform. Now that Github has made an arm64 macOS environment available to open source projects [1] we’re testing the platform along with the other Tier 1 supported platforms.

[1]
https://github.blog/changelog/2024-01-30-github-actions-introducing-the-new-m1-macos-runner-available-to-open-source/

For development of rustworkx the automated testing environment tooling used has switched from Tox to instead Nox. This is has no impact for end users and is only relevant if you contribute code to rustworkx.

0.14.0#

Prelude#

This is a new feature release of Rustworkx that adds many new features to the library. The highlights of this release are:

Fully type annotated for support with mypy and other tooling

Improvements to the graph coloring functions

This release supports running with Python 3.8 through 3.12. The minimum supported Rust version for building rustworkx and rustworkx-core from source is now 1.64.0. The minimum supported version of macOS for this release has been increased from 10.9 to 10.12. Also, the Linux ppc64le and s390x platform support has been downgraded from Tier 3 to Tier 4.

New Features#

Added two new random graph generator functions, directed_barabasi_albert_graph() and barabasi_albert_graph(), to generate a random graph using Barabási–Albert preferential attachment to extend an input graph. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

starting_graph = rustworkx.generators.path_graph(10)
random_graph = rustworkx.barabasi_albert_graph(20, 10, initial_graph=starting_graph)
mpl_draw(random_graph)

Added a new function to the rustworkx-core module rustworkx_core::generators barabasi_albert_graph() which is used to generate a random graph using Barabási–Albert preferential attachment to extend an input graph.

Added a new function all_shortest_paths() (and the graph type specific variants: graph_all_shortest_paths() and digraph_all_shortest_paths()) that finds every simple shortest path two nodes in a graph.

Added a new function to the rustworkx-core module rustworkx_core::shortest_path module all_shortest_path() which is used to find every simple shortest path in a graph.

Added a new function two_color to the rustworkx-core rustworkx_core::coloring module. This function is used to compute a two coloring of a graph and can also be used to determine if a graph is bipartite as it returns None when a two coloring is not possible.

Added a new function, two_color(), which is used to compute a two coloring for a graph. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.heavy_square_graph(5)
colors = rx.two_color(graph)
mpl_draw(graph, node_color=[colors[i] for i in range(len(graph))])

Added a new function, is_bipartite() to determine whether a given graph object is bipartite or not.

Added a new function, bridges() that finds the bridges of an undirected PyGraph. Bridges are edges that, if removed, would increase the number of connected components of a graph. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.PyGraph()
graph.extend_from_edge_list([
    (0, 1), (1, 2), (0, 2), (1, 3)
])
bridges = rustworkx.bridges(graph)
bridges_set = [set(edge) for edge in bridges]

colors = []
for edge in graph.edge_list():
  color = "red" if set(edge) in bridges_set else "black"
  colors.append(color)
mpl_draw(graph, edge_color=colors)

Added a new function bridges to the rustworkx_core:connectivity:biconnected module that finds the bridges of an undirected graph. Bridges are edges that, if removed, would increase the number of connected components of a graph. For example:

Added a new function, clear() that clears all nodes and edges from a PyGraph or PyDiGraph.

Added a new function, clear_edges() that clears all edges for PyGraph or rustworkx.PyDiGraph without modifying nodes.

Added method edge_indices_from_endpoints() which returns the indices of all edges between the specified endpoints. For PyDiGraph there is a corresponding method that returns the directed edges.

The PyGraph and the PyDiGraph classes have a new method filter_nodes() (or filter_nodes()). This method returns a NodeIndices object with the resulting nodes that fit some abstract criteria indicated by a filter function. For example:

from rustworkx import PyGraph

graph = PyGraph()
graph.add_nodes_from(list(range(5))) # Adds nodes from 0 to 5

def my_filter_function(node):
  return node > 2

indices = graph.filter_nodes(my_filter_function)
print(indices)

NodeIndices[3, 4]

The PyGraph and the PyDiGraph classes have a new method filter_edges() (or filter_edges()). This method returns a EdgeIndices object with the resulting edges that fit some abstract criteria indicated by a filter function. For example:

from rustworkx import PyGraph
from rustworkx.generators import complete_graph

graph = PyGraph()
graph.add_nodes_from(range(3))
graph.add_edges_from([(0, 1, 'A'), (0, 1, 'B'), (1, 2, 'C')])

def my_filter_function(edge):
  if edge:
    return edge == 'B'
  return False  
  
indices = graph.filter_edges(my_filter_function)
print(indices)

EdgeIndices[1]

Added a new algorithm function, rustworkx.floyd_warshall_successor_and_distance(), that calculates the shortest path distance and the successor nodes for all node pairs in PyGraph and PyDiGraph graphs.

Added a new function, graph_line_graph() to construct a line graph of a PyGraph object.

The line graph \(L(G)\) of a graph \(G\) represents the adjacencies between edges of G. \(L(G)\) contains a vertex for every edge in \(G\), and \(L(G)\) contains an edge between two vertices if the corresponding edges in \(G\) have a vertex in common.

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.PyGraph()
node_a = graph.add_node("a")
node_b = graph.add_node("b")
node_c = graph.add_node("c")
node_d = graph.add_node("d")
edge_ab = graph.add_edge(node_a, node_b, 1)
edge_ac = graph.add_edge(node_a, node_c, 1)
edge_bc = graph.add_edge(node_b, node_c, 1)
edge_ad = graph.add_edge(node_a, node_d, 1)

out_graph, out_edge_map = rx.graph_line_graph(graph)
assert out_graph.node_indices() == [0, 1, 2, 3]
assert out_graph.edge_list() == [(3, 1), (3, 0), (1, 0), (2, 0), (2, 1)]
assert out_edge_map == {edge_ab: 0, edge_ac: 1, edge_bc: 2, edge_ad: 3}
mpl_draw(out_graph, with_labels=True)

Added a new function, graph_greedy_edge_color() to color edges of a PyGraph object using a greedy approach.

This function works by greedily coloring the line graph of the given graph.

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.cycle_graph(7)
edge_colors = rx.graph_greedy_edge_color(graph)
assert edge_colors == {0: 0, 1: 1, 2: 0, 3: 1, 4: 0, 5: 1, 6: 2}
mpl_draw(graph, edge_color=[edge_colors[i] for i in range(graph.num_edges())])

Added a new function, graph_misra_gries_edge_color() to color edges of a PyGraph object using the Misra-Gries edge coloring algorithm.

The above algorithm is described in the paper paper: “A constructive proof of Vizing’s theorem” by Misra and Gries, 1992.

The coloring produces at most \(d + 1\) colors where \(d\) is the maximum degree of the graph.
import rustworkx as rx from rustworkx.visualization import mpl_draw graph = rx.generators.cycle_graph(7) edge_colors = rx.graph_misra_gries_edge_color(graph) assert edge_colors == {0: 0, 1: 1, 2: 2, 3: 0, 4: 1, 5: 0, 6: 2} mpl_draw(graph, edge_color=[edge_colors[i] for i in range(graph.num_edges())])

Added a new function, isolates(), which is used to find the isolates (nodes with a degree of 0) in a PyDiGraph or PyGraph.

Added a new function, isolates() to the rustworkx-core rustworkx_core::connectivity module which is used to find the isolates (nodes with a degree of 0).

Added method substitute_node_with_subgraph to the PyGraph class.

import rustworkx
from rustworkx.visualization import * # Needs matplotlib/

graph = rustworkx.generators.complete_graph(5)
sub_graph = rustworkx.generators.path_graph(3)

# Replace node 4 in this graph with sub_graph
# Make sure to connect the graphs at node 2 of the sub_graph
# This is done by passing a function that returns 2

graph.substitute_node_with_subgraph(4, sub_graph, lambda _, __, ___: 2)

# Draw the updated graph
mpl_draw(graph, with_labels=True)

Added a new function topological_generations() which stratifies a PyDiGraph into topological generations.

Added a new exception class GraphNotBipartite which is raised when a graph is not bipartite. The sole user of this exception is the graph_bipartite_edge_color() which will raise it when the user provided graph is not bipartite.

Added a new function, graph_bipartite_edge_color() to color edges of a PyGraph object. The function first checks whether a graph is bipartite, raising exception of type GraphNotBipartite if this is not the case. Otherwise, the function calls the algorithm for edge-coloring bipartite graphs, and returns a dictionary with key being the edge index and value being the assigned color.

The implemented algorithm is based on the paper “A simple algorithm for edge-coloring bipartite multigraphs” by Noga Alon, 2003.

The coloring produces at most \(d\) colors where \(d\) is the maximum degree of a node in the graph. The algorithm runs in time \(\mathcal{O}(n + m\log{}m)\), where \(n\) is the number of vertices and \(m\) is the number of edges in the graph.
import rustworkx as rx from rustworkx.visualization import mpl_draw graph = rx.generators.cycle_graph(8) edge_colors = rx.graph_bipartite_edge_color(graph) assert edge_colors == {0: 0, 1: 1, 2: 0, 3: 1, 4: 0, 5: 1, 6: 0, 7: 1} mpl_draw(graph, edge_color=[edge_colors[i] for i in range(graph.num_edges())])

Added two new random graph generator functions, directed_random_bipartite_graph() and undirected_random_bipartite_graph(), to generate a random bipartite graph. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

random_graph = rx.undirected_random_bipartite_graph(10, 5, 0.5, seed=20)
layout = rx.bipartite_layout(random_graph, set(range(10)))
mpl_draw(random_graph, pos=layout)

The functions graph_adjacency_matrix() and digraph_adjacency_matrix() now have the option to adjust parallel edge behavior. Instead of just the default sum behavior, the value in the output matrix can be the minimum (“min”), maximum (“max”), or average (“avg”) of the weights of the parallel edges. For example:

import rustworkx as rx
graph = rx.PyGraph()
a = graph.add_node("A")
b = graph.add_node("B")
c = graph.add_node("C")

graph.add_edges_from([
    (a, b, 3.0),
    (a, b, 1.0),
    (a, c, 2.0),
    (b, c, 7.0),
    (c, a, 1.0),
    (b, c, 2.0),
    (a, b, 4.0)
])

print("Adjacency Matrix with Summed Parallel Edges")
print(rx.graph_adjacency_matrix(graph, weight_fn= lambda x: float(x)))
print("Adjacency Matrix with Averaged Parallel Edges")
print(rx.graph_adjacency_matrix(graph, weight_fn= lambda x: float(x), parallel_edge="avg"))

Adjacency Matrix with Summed Parallel Edges
[[0. 8. 3.]
 [8. 0. 9.]
 [3. 9. 0.]]
Adjacency Matrix with Averaged Parallel Edges
[[0.         2.66666667 1.5       ]
 [2.66666667 0.         4.5       ]
 [1.5        4.5        0.        ]]

The rustworkx Python package is now fully typed with mypy. Building off of the previous 0.13.0 release which introduced partial type annotations to the library, rustworkx now includes type annotations for the entire public API.

Added a new exception class InvalidMapping which is raised when a function receives an invalid mapping. The sole user of this exception is the graph_token_swapper() which will raise it when the user provided mapping is not feasible on the provided graph.

Added has_path() which accepts as arguments a PyGraph or PyDiGraph and checks if there is a path from source to destination

from rustworkx import PyDiGraph, has_path

graph = PyDiGraph()
a = graph.add_node("A")
b = graph.add_node("B")
c = graph.add_node("C")
edge_list = [(a, b, 1), (b, c, 1)]
graph.add_edges_from(edge_list)

path_exists = has_path(graph, a, c)
assert(path_exists == True)

path_exists = has_path(graph, c, a)
assert(path_exists == False)

Added support for musl Linux platforms on x86_64 at Tier 3 and aarch64 at Tier 4.

Added a new keyword argument, preset_color_fn, to graph_greedy_color() which is used to provide preset colors for specific nodes when computing the graph coloring. You can optionally pass a callable to that argument which will be passed node index from the graph and is either expected to return an integer color to use for that node, or None to indicate there is no preset color for that node. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.generalized_petersen_graph(5, 2)

def preset_colors(node_index):
    if node_index == 0:
        return 3

coloring = rx.graph_greedy_color(graph, preset_color_fn=preset_colors)
colors = [coloring[node] for node in graph.node_indices()]

layout = rx.shell_layout(graph, nlist=[[0, 1, 2, 3, 4],[6, 7, 8, 9, 5]])
mpl_draw(graph, node_color=colors, pos=layout)

Added a new function greedy_node_color_with_preset_colors to the rustworkx-core module coloring. This new function is identical to the rustworkx_core::coloring::greedy_node_color except it has a second preset parameter which is passed a callable which is used to provide preset colors for particular node ids.

Added a new function, transitive_reduction() which returns the transitive reduction of a given PyDiGraph and a dictionary with the mapping of indices from the given graph to the returned graph. The given graph must be a Directed Acyclic Graph (DAG). For example:

from rustworkx import PyDiGraph
from rustworkx import transitive_reduction

graph = PyDiGraph()
a = graph.add_node("a")
b = graph.add_node("b")
c = graph.add_node("c")
d = graph.add_node("d")
e = graph.add_node("e")

graph.add_edges_from([
    (a, b, 1),
    (a, d, 1),
    (a, c, 1),
    (a, e, 1),
    (b, d, 1),
    (c, d, 1),
    (c, e, 1),
    (d, e, 1)
])

tr, _ = transitive_reduction(graph)
list(tr.edge_list())

[(0, 1), (0, 2), (1, 3), (2, 3), (3, 4)]

Ref: https://en.wikipedia.org/wiki/Transitive_reduction

Upgrade Notes#

The minimum supported rust version (MSRV) for rustworkx and rustworkx-core has been raised to 1.64.0. Previously you could build rustworkx or rustworkx-core using an MSRV of 1.56.1. This change was necessary as the upstream dependencies of rustworkx have adopted newer MSRVs.

The minimum required Python version was raised to Python 3.8. To use rustworkx, please ensure you are using Python >= 3.8.

The rustworkx function graph_token_swapper() now will raise an InvalidMapping exception instead of a PanicException when an invalid mapping is requested. This was done because a PanicException is difficult to catch by design as it is used to indicate an unhandled error. Using

The return type of the rustworkx-core function token_swapper() has been changed from Vec<(NodeIndex, NodeIndex)> to be Result<Vec<(NodeIndex, NodeIndex)>, MapNotPossible>. This change was necessary to return an expected error condition if a mapping is requested for a graph that is not possible. For example is if you have a disjoint graph and you’re trying to map nodes without any connectivity:
use rustworkx_core::token_swapper; use rustworkx_core::petgraph; let g = petgraph::graph::UnGraph::<(), ()>::from_edges(&[(0, 1), (2, 3) ]); let mapping = HashMap::from([ (NodeIndex::new(2), NodeIndex::new(0)), (NodeIndex::new(1), NodeIndex::new(1)), (NodeIndex::new(0), NodeIndex::new(2)), (NodeIndex::new(3), NodeIndex::new(3)), ]); token_swapper(&g, mapping, Some(10), Some(4), Some(50));
will now return Err(MapNotPossible) instead of panicking. If you were using this function before you’ll need to handle the result type.

Support for the Linux ppc64le pllatform has changed from tier 3 to tier 4 (as documented in Platform Support). This is a result of no longer being able to run tests during the pre-compiled wheel publishing jobs due to constraints in the available CI infrastructure. There hopefully shouldn’t be any meaningful impact resulting from this change, but as there are no longer tests being run to validate the binaries prior to publishing them there are no longer guarantees that the wheels for ppc64le are fully functional (although the likelihood they are is still high as it works on other platforms). If any issues are encountered with ppc64le Linux please open an issue.

For macOS the minimum version of macOS is now 10.12. Previously, the precompiled binary wheel packages for macOS x86_64 were published with support for >=10.9. However, because of changes in the support policy for the Rust programming language the minimum version needed to raised to macOS 10.12. If you’re using Qiskit on macOS 10.9 you can probably build Qiskit from source while the rustworkx MSRV (minimum supported Rust version) is < 1.74, but the precompiled binaries published to PyPI will only be compatible with macOS >= 10.12.

Support for the Linux s390x platform has changed from tier 3 to tier 4 (as documented in Platform Support). This is a result of no longer being able to run tests during the pre-compiled wheel publishing jobs due to constraints in the available CI infrastructure. There hopefully shouldn’t be any meaningful impact resulting from this change, but as there are no longer tests being run to validate the binaries prior to publishing them there are no longer guarantees that the wheels for s390x are fully functional (although the likelihood they are is still high as it works on other platforms). If any issues are encountered with s390x Linux please open an issue.

Deprecation Notes#

The legacy retworkx package that operates as a backwards compatibility alias for rustworkx has been marked as deprecated. If you’re using the retworkx package it will now emit a DeprecationWarning on import.

Bug Fixes#

Fixed the behavior of graph_all_simple_paths() and digraph_all_simple_paths() when min_depth is set to 0. Refer to #955 for more information.

Fixed an issue where the directed_gnp_random_graph() and the gnp_random_graph() for directed graphs produced a graph where lower node numbers had only a small number of edges compared to what was expected.

Other Notes#

This version of rustworkx is explicitly pinned to the Numpy 1.x series, because it includes compiled extensions that are not yet compiled against the as-yet-unreleased Numpy 2.x series. We will release a new version of rustworkx with Numpy 2.x support as soon as feasible.

We cannot prevent your package manager from resolving to older versions of rustworkx (which do not have the same pin but are still likely to be incompatible) if you forcibly try to install rustworkx alongside Numpy 2, before we have released a compatible version.

0.13.0#

Prelude#

This release is major feature release of Rustworkx that adds some new features to the library. The highlights of this release are:

An expansion of the functions exposed by rustworkx-core to including a new graph generator module.

New link analysis functions such as page rank

Expanded centrality measure functions

Added partial type annotations to the library including for the PyDiGraph and PyGraph classes. This enables type checking with mypy

This is also the final rustworkx release that supports running with Python 3.7. Starting in the 0.14.0 release Python >= 3.8 will be required to use rustworkx. This release also increased the minimum supported Rust version for compiling rustworkx and rustworkx-core from source to 1.56.1.

New Features#

Added a new method, make_symmetric(), to the PyDiGraph class. This method is used to make all the edges in the graph symmetric (there is a reverse edge in the graph for each edge). For example:

import rustworkx as rx
from rustworkx.visualization import graphviz_draw

graph = rx.generators.directed_path_graph(5, bidirectional=False)
graph.make_symmetric()
graphviz_draw(graph)

Added a new function, edge_betweenness_centrality() to compute edge betweenness centrality of all edges in a PyGraph or PyDiGraph object. The algorithm used in this function is based on: Ulrik Brandes, On Variants of Shortest-Path Betweenness Centrality and their Generic Computation. Social Networks 30(2):136-145, 2008. Edge betweenness centrality of an edge \(e\) is the sum of the fraction of all-pairs shortest paths that pass through \(e\)

\[c_B(e) =\sum_{s,t \in V} \frac{\sigma(s, t|e)}{\sigma(s, t)}\]

where \(V\) is the set of nodes, \(\sigma(s, t)\) is the number of shortest \((s, t)\)-paths, and \(\sigma(s, t|e)\) is the number of those paths passing through edge \(e\). For example, the following computes the edge betweenness centrality for all edges in a 5x5 grid graph and uses the result to color the edges in a graph visualization:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.generators.grid_graph(5, 5) btw = rustworkx.edge_betweenness_centrality(graph) # Color edges in graph visualization with edge betweenness centrality colors = [] for i in graph.edge_indices(): colors.append(btw[i]) mpl_draw(graph, edge_color=colors)

Added a new function to rustworkx-core edge_betweenness_centrality to the rustworkx_core:centrality module which computes the edge betweenness centrality of all edges in a given graph.

Two new functions, find_cycle and cycle_basis, have been added to the rustworkx-core crate in the connectivity module. These functions can be used to find a cycle in a petgraph graph or to find the cycle basis of a graph.

Added a new function, hits() which is used to compute the hubs and authorities for all nodes in a given directed graph. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.directed_hexagonal_lattice_graph(2, 2)
hubs, _ = rx.hits(graph)

# Generate a color list
colors = []
for node in graph.node_indices():
    hub_score = hubs[node]
    graph[node] = hub_score
    colors.append(hub_score)
mpl_draw(
    graph,
    with_labels=True,
    node_color=colors,
    node_size=650,
    labels=lambda x: "{0:.2f}".format(x)
)

Added a new function, katz_centrality() which is used to compute the Katz centrality for all nodes in a given graph. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.hexagonal_lattice_graph(4, 4)
centrality = rx.katz_centrality(graph)

# Generate a color list
colors = []
for node in graph.node_indices():
    centrality_score = centrality[node]
    graph[node] = centrality_score
    colors.append(centrality_score)
mpl_draw(
    graph,
    with_labels=True,
    node_color=colors,
    node_size=650,
    labels=lambda x: "{0:.2f}".format(x)
)

Added a new function to rustworkx-core katz_centrality to the rustworkx_core::centrality modules which is used to compute the Katz centrality for all nodes in a given graph.

Added a new function, longest_simple_path() which is used to search all the simple paths between all pairs of nodes in a graph and return the longest path found. For example:

import rustworkx as rx

graph = rx.generators.binomial_tree_graph(5)
longest_path = rx.longest_simple_path(graph)
print(longest_path)

NodeIndices[31, 30, 28, 24, 16, 0, 8, 12, 14, 15]

Then visualizing the nodes in the longest path found:

from rustworkx.visualization import mpl_draw

path_set = set(longest_path)
colors = []
for index in range(len(graph)):
    if index in path_set:
      colors.append('r')
    else:
      colors.append('#1f78b4')
mpl_draw(graph, node_color=colors)

Added a new function longest_simple_path_multiple_targets() to rustworkx-core. This function will return the longest simple path from a source node to a HashSet of target nodes.

Added a new function, pagerank() which is used to compute the PageRank score for all nodes in a given directed graph. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.directed_hexagonal_lattice_graph(2, 2)
ranks = rx.pagerank(graph)

# Generate a color list
colors = []
for node in graph.node_indices():
    pagerank_score = ranks[node]
    graph[node] = pagerank_score
    colors.append(pagerank_score)
mpl_draw(
    graph,
    with_labels=True,
    node_color=colors,
    node_size=650,
    labels=lambda x: "{0:.2f}".format(x)
)

Three new random graph generators, gnp_random_graph, gnm_random_graph and random_geometric_graph, have been added to the rustworkx-core crate in the generators module. The gnp_random_graph takes inputs of the number of nodes and a probability for adding edges. The gnp_random_graph takes inputs of the number of nodes and number of edges. The random_geometric_graph creates a random graph within an n-dimensional cube.

Added a new function, bfs_predecessors(), which is used to return a list of predecessors in a reversed bread-first traversal from a specified node. This is analogous to the existing bfs_successors() method.

Add a method find_predecessor_node_by_edge() to get the immediate predecessor of a node which is connected by the specified edge.

Added a new function, graph_token_swapper(), which performs an approximately optimal token swapping algorithm based on:

Approximation and Hardness for Token Swapping by Miltzow et al. (2016) https://arxiv.org/abs/1602.05150

that supports partial mappings (i.e. not-permutations) for graphs with missing tokens.

Added a new function token_swapper() to the new rustworkx-core module rustworkx_core::token_swapper. This function performs an approximately optimal token swapping algorithm based on:

Approximation and Hardness for Token Swapping by Miltzow et al. (2016) https://arxiv.org/abs/1602.05150

that supports partial mappings (i.e. not-permutations) for graphs with missing tokens.

Added a new function, closeness_centrality() to compute the closeness centrality of all nodes in a PyGraph or PyDiGraph object.

The closeness centrality of a node \(u\) is defined as the the reciprocal of the average shortest path distance to \(u\) over all \(n-1\) reachable nodes. In it’s general form this can be expressed as:

\[C(u) = \frac{n - 1}{\sum_{v=1}^{n-1} d(v, u)},\]

where \(d(v, u)\) is the shortest-path distance between \(v\) and \(u\), and \(n\) is the number of nodes that can reach \(u\). For example, to visualize the closeness centrality of a graph:

import matplotlib.pyplot as plt

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.hexagonal_lattice_graph(4, 4)
centrality = rx.closeness_centrality(graph)
# Generate a color list
colors = []
for node in graph.node_indices():
    colors.append(centrality[node])
# Generate a visualization with a colorbar
plt.rcParams['figure.figsize'] = [15, 10]
ax = plt.gca()
sm = plt.cm.ScalarMappable(norm=plt.Normalize(
    vmin=min(centrality.values()),
    vmax=max(centrality.values())
))
plt.colorbar(sm, ax=ax)
plt.title("Closeness Centrality of a 4 x 4 Hexagonal Lattice Graph")
mpl_draw(graph, node_color=colors, ax=ax)

Added new generator functions, empty_graph(), and directed_empty_graph() to the rustworkx.generators module that will generate an empty graph. For example:
import rustworkx.generators from rustworkx.visualization import mpl_draw graph = rustworkx.generators.empty_graph(4) mpl_draw(graph)

Added new generator functions, complete_graph(), and directed_complete_graph() to the rustworkx.generators module that will generate a complete graph. These functions are equivalent to calling the mesh_graph() and directed_mesh_graph() functions. For example:
import rustworkx.generators from rustworkx.visualization import mpl_draw graph = rustworkx.generators.complete_graph(4) mpl_draw(graph)

Added a new generators module to the rustworkx-core crate. This module contains functions for generating graphs. These functions are generic on the output graph type and can be used to create graph objects for any type that implement the required petgraph traits.

Added partial type annotations to the library, including for the PyDiGraph and PyGraph classes. This enables statically type checking with mypy.

Added a new function, greedy_node_color, to rustworkx-core in a new coloring module. It colors a graph using a greedy graph coloring algorithm.

The function core_number has been added to the rustworkx-core crate in the connectivity module. It computes the k-core number for the nodes in a graph.

Upgrade Notes#

Passing a negative value to the probability argument to the gnp_directed_random_graph() or the gnp_undirected_random_graph() function will now cause an OverflowError to be raised. Previously, a ValueError would be raised in this situation. This was changed to be consistent with other similar error conditions in other functions in the library.

The minimum supported Rust version has been increased from 1.48 to 1.56.1. This applies to both building the rustworkx package from source as well as the rustworkx-core crate. This change was made to facilitate using newer versions of our upstream dependencies as well as leveraging newer Rust language features.

Bug Fixes#

Fixed the check_cycle attribute not being preserved when copying PyDiGraph with copy.copy() and copy.deepcopy(). Fixed #836

Fixed an issue when using copy.deepcopy() on PyDiGraph and PyGraph objects when there were removed edges from the graph object. Previously, if there were any holes in the edge indices caused by the removal the output copy of the graph object would incorrectly have flatten the indices. This has been corrected so that the edge indices are recreated exactly after a deepcopy(). Fixed #585

Fixed a compatibility issue when building rustworkx-core with priority-queue 1.3.0. Fixed #744

Fixed an issue with several PyDiGraph and PyGraph methods that removed nodes where previously when calling these methods the PyDiGraph.node_removed attribute would not be updated to reflect that nodes were removed.

Fixed an issue with the custom sequence return types, BFSSuccessors, NodeIndices, EdgeList, WeightedEdgeList, EdgeIndices, and Chains where they previously were missing certain attributes that prevented them being used as a sequence for certain built-in functions such as reversed(). Fixed #696.

rustworkx.PyGraph.add_edge() and rustworkx.PyDiGraph.add_edge() and now raises an IndexError when one of the nodes does not exist in the graph. Previously, it caused the Python interpreter to exit with a PanicException

0.12.0#

Prelude#

This release introduces some major changes to the Rustworkx (formerly retworkx) project. The first change is the library has been renamed from retworkx to rustworkx, and the retworkx-core rust crate has been renamed rustworkx-core. This was done out of respect for a request from the maintainers of the NetworkX library. For the current release the retworkx library will still continue to work as it has without any notification, but starting in the 0.13.0 release a DeprecationWarning will be emitted when importing from retworkx and in the 1.0.0 release we will drop support for the legacy name. For the retworkx-core crate, there will no longer be any releases under that name on crates.io and all future versions of the library will be released as rustworkx-core.

Additionally this release adds support for Python 3.11 and also moves to manylinux2014 for all precompiled Linux binaries we publish to PyPI. The minimum supported Rust version for building rustworkx from source has increased to Rust 1.48.

This release also includes several new features, some highlights are:

Support for graph attributes under the attrs attribute

New serialization format support (a graphml parser, read_graphml(), and a node link JSON generator, node_link_json())

New algorithms functions including:

Eigenvector Centrality

Stoer–Wagner Min-Cut algorithm

Bellman-Ford shortest path algorithm

New Features#

Added a new function, eigenvector_centrality() which is used to compute the eigenvector centrality for all nodes in a given graph. For example:

import rustworkx as rx
from rustworkx.visualization import mpl_draw

graph = rx.generators.hexagonal_lattice_graph(4, 4)
centrality = rx.eigenvector_centrality(graph)

# Generate a color list
colors = []
for node in graph.node_indices():
    centrality_score = centrality[node]
    graph[node] = centrality_score
    colors.append(centrality_score)
mpl_draw(
    graph,
    with_labels=True,
    node_color=colors,
    node_size=650,
    labels=lambda x: "{0:.2f}".format(x)
)

Added a new function to rustworkx-core eigenvector_centrality to the rustworkx_core::centrality modules which is used to compute the eigenvector centrality for all nodes in a given graph.

Added a new keyword arguments, index_output, to the layers() function. When set to True the output of the function is a list of layers as node indices. The default output is still a list of layers of node data payloads as before.

Added a new function, node_link_json(), which is used to generate JSON node-link data representation of an input PyGraph or PyDiGraph object. For example, running:

import retworkx

graph = retworkx.generators.path_graph(weights=['a', 'b', 'c'])
print(retworkx.node_link_json(graph, node_attrs=lambda n: {'label': n}))

will output a JSON payload equivalent (identical except for whitespace) to:

{
  "directed": false,
  "multigraph": true,
  "attrs": null,
  "nodes": [
    {
      "id": 0,
      "data": {
        "label": "a"
      }
    },
    {
      "id": 1,
      "data": {
        "label": "b"
      }
    },
    {
      "id": 2,
      "data": {
        "label": "c"
      }
    }
  ],
  "links": [
    {
      "source": 0,
      "target": 1,
      "id": 0,
      "data": null
    },
    {
      "source": 1,
      "target": 2,
      "id": 1,
      "data": null
    }
  ]
}

Added a new algorithm function, rustworkx.stoer_wagner_min_cut() that uses the Stoer Wagner algorithm for computing a weighted minimum cut in an undirected PyGraph. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.grid_graph(2, 2)
cut_val, partition = rustworkx.stoer_wagner_min_cut(graph)

colors = [
    'orange' if node in partition else 'blue' for node in graph.node_indexes()
]
mpl_draw(graph, node_color=colors)

Add two new functions which calculates the tensor product of two graphs graph_tensor_product() for undirected graphs and digraph_tensor_product() for directed graphs. For example:

import rustworkx
from rustworkx.visualization import mpl_draw
graph_1 = rustworkx.generators.path_graph(2)
graph_2 = rustworkx.generators.path_graph(3)
graph_product, _ = rustworkx.graph_tensor_product(graph_1, graph_2)

mpl_draw(graph_product)

Added new functions to compute the all-pairs shortest path in graphs with negative edge weights using the Bellman-Ford algorithm with the SPFA heuristic:
- rustworkx.all_pairs_bellman_ford_path_lengths()
- rustworkx.all_pairs_bellman_ford_shortest_paths()

Added a new function all_pairs_all_simple_paths() which is used to return all simple paths between all pairs of nodes in a graph. It can also be used with a optional min_depth and cutoff parameters to filter the results based on path lengths. For example:
```
from rustworkx.generators import grid_graph
from rustworkx import all_pairs_all_simple_paths

g = grid_graph(2, 3)
paths = all_pairs_all_simple_paths(g, min_depth=3, cutoff=3)
```
will return a dictionary of dictionaries where the 2 dictionary keys are the node indices and the inner value is the list of all simple paths of length 3 between those 2 nodes.

The rustworkx-core rustworkx_core::connectivity module has a new function all_simple_paths_multiple_targets this is similar to the all_simple_paths() method in petgraph’s algo module but instead of returning an iterator that will yield the all the simple path from a source to the target node it instead will build a DictMap of all the simple paths from a source node to all targets in a provided HashSet of target node indices.

Added new functions to compute negative cycles and the shortest path in graphs with negative edge weights using the Bellman-Ford algorithm with the SPFA heuristic:

Added a concept of graph attributes to the PyDiGraph and PyGraph classes. The attributes are accessible via the attrs attribute of the graph objects and can be modified in place. Additionally, they can be set initially when creating the object via the constructor. For example:
```
import rustworkx as rx

graph = rx.PyGraph(attrs=dict(day="Friday"))
graph.attrs['day'] = "Monday"
```
The attributes can contain any Python object, not just a dictionary. For example:
```
class Day:

    def __init__(self, day):
        self.day = day

graph = rx.PyGraph(attrs=Day("Friday"))
graph.attrs = Day("Monday")
```
If attrs is not set it will default to None.

The PyGraph.subgraph() and PyDiGraph.subgraph() methods have a new keyword argument preserve_attributes which can be set to True to copy by reference the contents of the attrs attribute from the graph to the subgraph’s attrs attribute.

Implements a new function is_planar() that checks whether an undirected PyGraph is planar.

import rustworkx as rx

graph = rx.generators.mesh_graph(5)
print('Is K_5 graph planar?', rx.is_planar(graph))

Is K_5 graph planar? False

The rustworkx-core connectivity module has 3 new functions, connected_components, number_connected_components, and bfs_undirected. These functions are based on the existing connected_components(), number_connected_components(), and bfs_undirected() in rustworkx.

Added a new function read_graphml() that generates a rustworkx graph object (a PyGraph or a PyDiGraph) from a file written in GraphML format. GraphML is an xml serialization format for representing graph files.

Added a new function, simple_cycles(), which is an implementation of Johnson’s algorithm for finding all elementary cycles in a directed graph.

Upgrade Notes#

The return type for the PyGraph method add_edges_from() and add_edges_from_no_data() has changed from a list of integer edge indices to an EdgeIndices object. The EdgeIndices class is a read-only sequence type of integer edge indices. For the most part this should be fully compatible except if you were mutating the output list or were explicitly type checking the return. In these cases you can simply cast the EdgeIndices object with list().

This release no longer provides binaries that support the manylinux2010 packaging specification. All the precompiled binaries for Linux platforms are built against manylinux2014. This change is required due to changes in the GLIBC versions supported by the latest versions of the Rust compiler in addition to the manylinux2010 platform no longer being supported. If you need to run Rustworkx on a platform only compatible with manylinux2010 starting with this release you will need to build and install from source (which includes the sdist published to PyPI, so pip install rustworkx will continue to work assuming you have a Rust compiler installed) and also use a Rust compiler with a version < 1.64.0.

The minimum supported Rust version for building rustworkx has been raised from 1.41 to 1.48. To compile rustworkx from source you will need to ensure you have at Rustc >=1.48 installed.

The minimum supported Python version for using rustworkx has been raised to Python 3.7. To use rustworkx you will need to ensure you are using Python >=3.7.

The retworkx package has been renamed to rustworkx. This was done out of respect for a request from the maintainers of the NetworkX library. For the time being the retworkx name will continue to work, however any package requirements or imports using retworkx should be renamed to rustworkx.

Bug Fixes#

The custom sequence return classes:
- BFSSSuccessors
- NodeIndices
- EdgeList
- WeightedEdgeList
- EdgeIndices
- Chains
now correctly handle slice inputs to __getitem__. Previously if you tried to access a slice from one of these objects it would raise a TypeError. For example, if you had a NodeIndices object named nodes containing [0, 1, 3, 4, 5] if you did something like:
```
nodes[0:3]
```
it would return a new NodeIndices object containing [0, 1, 3] Fixed #590

0.11.0#

Prelude#

This release includes many new features and bug fixes. Highlights include additional methods to improve working with edges in both PyGraph and PyDiGraph, several new graph generators, and a new set of interactive traversal functions: bfs_search(), dfs_search(), dijkstra_search(), and TopologicalSorter, which enable iterative interaction with the graph during different types of traversal.

This release also introduces a new separate Rust crate rustworkx-core which is a library for use in Rust that’s built on top of petgraph that extends the functionality provided by petgraph. The functionality in this crate is generic and can work with any petgraph graph, not just the PyGraph amd PyDiGraph.

The 0.11.0 release fully supports Python 3.10. Precompiled binaries for Python 3.10 are published on PyPI (previous releases worked with 3.10 but required compiling from source to install). This is also the last release to support Python 3.6. Starting in rustworkx 0.12.0, Python >=3.7 will be required to run rustworkx. Additionally, for users compiling rustworkx from source, this will be the last release with a minimum supported Rust version (MSRV) of 1.41. In rustworkx 0.12.0, the MSRV will be increased to 1.48.

New Features#

Added a new function, cartesian_product() (and its per type variants digraph_cartesian_product() and graph_cartesian_product()), which calculates the Cartesian product of two graphs. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph_1 = rustworkx.generators.path_graph(2)
graph_2 = rustworkx.generators.path_graph(3)

graph_product, _ = rustworkx.cartesian_product(graph_1, graph_2)

mpl_draw(graph_product)

Added a new method, node_indices(), to the PyDiGraph and PyGraph classes. This method is identical to the previously existing node_indexes() method but changes the name to be consistent with the use of “indices” throughout the rest of rustworkx. The node_indexes() will likely be deprecated in a future release prior to it’s removal in an eventual 1.0 release.

The unweighted_average_shortest_path_length() function has a new kwarg disconnected. When disconnected is set to True the output value calculated by the function will only account for connected node pairs.

Added a new generator function, barbell_graph(), to the rustworkx.generators module that will generate a barbell graph. For example:

import rustworkx.generators
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.barbell_graph(4, 3)
mpl_draw(graph)

Added a new bfs_search() (and it’s per type variants graph_bfs_search() and digraph_bfs_search()) that traverses the graph in a breadth-first manner and emits events at specified points. The events are handled by a visitor object that subclasses BFSVisitor through the appropriate callback functions. For example:

import rustworkx
from rustworkx.visit import BFSVisitor


class TreeEdgesRecorder(BFSVisitor):

    def __init__(self):
        self.edges = []

    def tree_edge(self, edge):
        self.edges.append(edge)

graph = rustworkx.PyGraph()
graph.extend_from_edge_list([(1, 3), (0, 1), (2, 1), (0, 2)])
vis = TreeEdgesRecorder()
rustworkx.bfs_search(graph, [0], vis)
print('Tree edges:', vis.edges)

Tree edges: [(0, 2, None), (0, 1, None), (1, 3, None)]

Added a new function articulation_points() that finds the articulation points of an undirected PyGraph. An articulation point or cut vertex is any node whose removal increases the number of connected components of a graph. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.PyGraph()
graph.extend_from_edge_list([
    (0, 1), (1, 2), (0, 2), (1, 3)
])
points = rustworkx.articulation_points(graph)

colors = ['black'] * len(graph)
for node in points:
    colors[node] = 'blue'

mpl_draw(graph, node_color=colors)

Added a new function biconnected_components() that returns the biconnected components of an undirected PyGraph. A biconnected component is a maximal subgraph that remains connected after removal of a node. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.PyGraph()
graph.extend_from_edge_list([
    (0, 1), (1, 2), (0, 2), (1, 3), (2, 4)
])
components = rustworkx.biconnected_components(graph)

COLORS = ["blue", "red", "orange"]
edge_colors = []
for (u, v) in graph.edge_list():
  if (u, v) in components:
      comp = components[(u, v)]
  else:
      comp = components[(v, u)]
  edge_colors += [COLORS[comp]] 

mpl_draw(graph, node_color='black', node_size=150,
        edge_color=edge_colors)

Added a new function chain_decomposition() that finds a chain decomposition of an undirected PyGraph. A chain decomposition is a set of cycles or paths derived from the set of fundamental cycles of a depth-first tree. It’s defined in https://doi.org/10.1016/j.ipl.2013.01.016 For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.PyGraph()
graph.extend_from_edge_list([
    (0, 1), (0, 2), (1, 2), (2, 3),
    (3, 4), (3, 5), (4, 5),
])
chains = rustworkx.chain_decomposition(graph)

def color_edges(graph, chains): 
    COLORS = ['blue', 'red']

    edges_in_chain = {}
    for idx, chain in enumerate(chains):
        for edge in chain:
            edge = tuple(sorted(edge))
            edges_in_chain[edge] = COLORS[idx]
        
    edge_colors = []
    for edge in graph.edge_list():
        edge = tuple(sorted(edge))
        edge_colors += [edges_in_chain.get(edge, 'black')]

    return edge_colors

mpl_draw(graph, node_color='black', node_size=150,
        edge_color=color_edges(graph, chains))

Added new constructor methods rustworkx.PyDiGraph.from_complex_adjacency_matrix() and rustworkx.PyGraph.from_complex_adjacency_matrix() for creating a PyDiGraph and a PyGraph respectively from a numpy adjacency matrix with a complex dtype. For example:

import numpy as np

import rustworkx
from rustworkx.visualization import mpl_draw

adj_matrix = np.array([
    [0, 0 - 1j, 0, 1 + 2j],
    [0 - 1j, 0, 1 + 0j, 0],
    [0, 1 + 0j, 0, 2 + 0.5j],
    [1 + 2j, 0, 2 + 0.5j, 0]
], dtype=complex)
graph = rustworkx.PyDiGraph.from_complex_adjacency_matrix(adj_matrix)

mpl_draw(graph, with_labels=True, edge_labels=str)

Adds connected_components() for finding connected components in an undirected PyGraph graph.

Adds number_connected_components() for finding the number of connected components in an undirected PyGraph graph.

Adds node_connected_component() for finding the connected component that a node belongs in an undirected PyGraph graph.

Adds is_connected() for checking if an undirected PyGraph graph is connected.

Added new graph methods rustworkx.PyDiGraph.contract_nodes(), and rustworkx.PyGraph.contract_nodes(). These methods can be used to replace a set of graph nodes with a single new equivalent node. Incoming edges and outgoing edges of and to the replaced set become the incoming and outgoing edges of the new node, respectively. In a multi-graph, all edges are preserved by default. For all graph types, parallel edges can optionally be combined via a user-specified Python callable. rustworkx.PyDiGraph.contract_nodes() supports cycle checking / guarding to prevent the contraction from introducing cycles. In the following example, two nodes are contracted to a single new node. First, creating a new graph:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.generators.directed_path_graph(5) mpl_draw(graph, with_labels=True)
perform the contraction:
graph.contract_nodes([2, 3], "abc") mpl_draw(graph, with_labels=True)

Added a new dfs_search() (and it’s per type variants graph_dfs_search() and digraph_dfs_search()) that traverses the graph in a depth-first manner and emits events at specified points. The events are handled by a visitor object that subclasses DFSVisitor through the appropriate callback functions. For example:

import rustworkx
from rustworkx.visit import DFSVisitor

class TreeEdgesRecorder(DFSVisitor):

    def __init__(self):
        self.edges = []

    def tree_edge(self, edge):
        self.edges.append(edge)

graph = rustworkx.PyGraph()
graph.extend_from_edge_list([(1, 3), (0, 1), (2, 1), (0, 2)])
vis = TreeEdgesRecorder()
rustworkx.dfs_search(graph, [0], vis)
print('Tree edges:', vis.edges)

Tree edges: [(0, 2, None), (2, 1, None), (1, 3, None)]

Added a new dijkstra_search() (and it’s per type variants graph_dijkstra_search() and digraph_dijkstra_search()) that traverses the graph using dijkstra algorithm and emits events at specified points. The events are handled by a visitor object that subclasses DijkstraVisitor through the appropriate callback functions. For example:

import rustworkx
from rustworkx.visit import DijkstraVisitor


class DijkstraTreeEdgesRecorder(rustworkx.visit.DijkstraVisitor):

    def __init__(self):
        self.edges = []
        self.parents = dict()

    def discover_vertex(self, v, _):
        u = self.parents.get(v, None)
        if u is not None:
            self.edges.append((u, v))

    def edge_relaxed(self, edge):
        u, v, _ = edge
        self.parents[v] = u

graph = rustworkx.PyGraph()
graph.extend_from_weighted_edge_list([(1, 3, 1), (0, 1, 10), (2, 1, 1), (0, 2, 1)])
vis = DijkstraTreeEdgesRecorder()
rustworkx.graph_dijkstra_search(graph, [0], float, vis)
print('Tree edges:', vis.edges)

Tree edges: [(0, 2), (2, 1), (1, 3)]

Added a new method, incident_edges(), to the PyGraph and PyDiGraph class. This method returns a list of edge indices for edges incident to a provided node.

Added a new method, incident_edge_index_map(), to the PyGraph and PyDiGraph class. This method returns a mapping of edge indices for edges incident to a provided node to the endpoint and weight tuple for that edge index. For example:
import rustworkx graph = rustworkx.PyGraph() graph.extend_from_weighted_edge_list([(0, 1, "A"), (0, 2, "B")]) print(graph.incident_edge_index_map(0))
EdgeIndexMap{1: (0, 2, B), 0: (0, 1, A)}

Added a new method, get_edge_data_by_index(), to the PyGraph and PyDiGraph classes. This method returns the data payload for an edge in the graph from its index.

Added a new method, get_edge_endpoints_by_index(), to the PyGraph and PyDiGraph classes. This method returns the edge’s endpoint tuple for an edge in the graph from its index.

Added two new methods, out_edges() and in_edges() to the PyGraph class. These methods are the duals of the PyDiGraph methods, out_edges() and in_edges() and return a WeightedEdgeList of the incident edges for a node.

The algorithm used for the steiner_tree() implementation has been replaced by a faster algorithm based on: https://doi.org/10.1016/0020-0190(88)90066-X. This new implementation achieves the same approximation ratio as the algorithm used previously, so there should be no change in functionality.

Added a new function, full_rary_tree() that adds support for generating a full \(r\)-ary tree of \(n\) nodes. For example:

import rustworkx.generators
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.full_rary_tree(3, 5)
mpl_draw(graph)

Added a new function, lollipop_graph() that adds support for generating lollipop graphs. For example:

import rustworkx.generators
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.lollipop_graph(4, 3)
mpl_draw(graph)

Added a new class, TopologicalSorter, which provides functionality to topologically sort a directed graph. It gives the ability to process nodes while they are becoming ready and then mark them as done, so more nodes can be freed.

The betweenness_centrality() (and it’s per type variants graph_betweenness_centrality() and digraph_betweenness_centrality()) is now multithreaded. For larger graphs this can significantly improve the runtime performance of the function. By default any graphs with < 50 nodes will still execute in a single thread, while larger graphs will be executed in parallel. The size of the graph to start running in parallel can be adjusted using the new parallel_threshold kwarg. Additionally, the environment variable RAYON_NUM_THREADS can be used how many threads will be used when run in parallel. By default it will use a thread for each CPU on the local system.

Added a new function, generalized_petersen_graph() that adds support for generating generalized Petersen graphs. For example:

import rustworkx.generators
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.generalized_petersen_graph(5, 2)
layout = rustworkx.shell_layout(graph, nlist=[[0, 1, 2, 3, 4], [6, 7, 8, 9, 5]])
mpl_draw(graph, pos=layout)

Added a new workspace crate, rustworkx-core as part of rustworkx. This is a standalone Rust library that is built on top of petgraph that provides general algorithms and graph functionality that rustworkx needs. This new crate only exposes a Rust interface that is general for any petgraph graph and can be used by any downstream Rust project that wants the extra functionality that rustworkx exposes, but without Python.

It’s also worth noting as this is the first release of rustworkx-core there may be breaking API changes in a subsequent release. While we will attempt to follow the standard deprecation and stability policy, since we’re not necessarily fully committed to the current API and without having a user outside of rustworkx, there may be gaps or issues which require breaking changes.

A new kwarg, keep_attributes, has been added to the NetworkX graph converter function networkx_converter(). When this argument is set to True the node attributes from the input NetworkX graph will be preserved. The data payload for the output rustworkx graph will be a dictionary containing the attributes, with an extra "__networkx_node__" key containing the node from NetworkX. For example:

import networkx as nx
import rustworkx as rx

g = nx.Graph()
g.add_nodes_from([
  ("A", {"color": "turquoise", "size": "extra large"}),
  ("B", {"color": "fuchsia", "size": "tiny"}),
])
g.add_edge("A", "B")
rx_graph = rx.networkx_converter(g, keep_attributes=True)
print(rx_graph.nodes())
print(rx_graph.weighted_edge_list())

will output:

[{'color': 'turquoise', 'size': 'extra large', '__networkx_node__': 'A'}, {'color': 'fuchsia', 'size': 'tiny', '__networkx_node__': 'B'}]
WeightedEdgeList[(0, 1, {})]

Upgrade Notes#

The default behavior for how the unweighted_average_shortest_path_length() function handles disconnected graphs has been changed. Previously, disconnected pairs of nodes was assumed to have zero distance which is arguably incorrect/unexpected behavior. To make this more consistent with user expectations this has been changed to an infinite value. In addition, an extra kwarg disconnected was added where, if set to True, the average is taken only over connected node pairs. By default, it’s set to False. If the previous behavior of treating disconnected pairs of nodes as having a distance of zero is desired, it can be reproduced using the rest of rustworkx API like:
```
import rustworkx

graph = rustworkx.undirected_gnm_random_graph(20, 10, seed=42)
n = len(graph)
d_mat = rustworkx.distance_matrix(graph, null_value=0.0)
avg_shortest_path = d_mat.sum() / (n * (n - 1.0))
```

The optional dependencies for graphviz_draw() (as documented by the graphviz optional extra, which can be installed via pip install rustworkx[graphviz]) no longer requires the pydot library. Only the pillow library is needed now to generate visualizations using graphviz.

Bug Fixes#

Fixed an issue with the binomial_tree_graph() and directed_binomial_tree_graph() generator functions in rustworkx.generators where passing an order value >= 60 would cause an overflow and raise a PanicException caused by the internal Rust panic when overflowing or exceeding the max Vec size. Instead the function will raise an OverflowError and indicate the specified order is too large. Fixed #457

Fixed an issue where distance_matrix() and k_shortest_path_lengths() would previously panic if the input graph contains holes in the node indices.

Fixed the rustworkx.PyGraph.degree() method when running on a node that has self-loops. Previously, the method would increment the node’s degree of a self-loop by one instead of two. Fixed #517.

Fixed the dfs_edges() function to have the source argument default to None as was documented. Previously, the source argument was incorrectly required and had no default value. Fixed #515.

Fixed an oversight in the union() function where user-defined values for the merge_nodes and merge_edges arguments were being ignored.

Fixed an oversight in shortest path algorithms functions, such as: dijkstra_shortest_paths(), dijkstra_shortest_path_lengths(), all_pairs_dijkstra_path_lengths(), all_pairs_dijkstra_shortest_paths(), astar_shortest_path(), and k_shortest_path_lengths() which would previously incorrectly accept edge weights with negative or NaN values. Fixed #525.

Fixed an issue with the heavy_hex_graph(), directed_heavy_hex_graph(), heavy_square_graph(), and directed_heavy_square_graph() generator functions. When the input parameter d was set to 1 these functions would previously raise a pyo3_runtime.PanicException instead of returning the expected graph (a single node). Fixed #452

0.10.2#

Prelude#

This release is a bugfix release that fixes some issues found since the 0.10.1 release. The fixes in this release are primarily in the vf2 implementation and most importantly fixes the output of vf2_mapping() with an empty input graph and the output of vf2_mapping(), is_isomorphic(), and is_subgraph_isomorphic() with graphs that have parallel edges.

New Features#

Add a new function graph_union() that returns the union of two PyGraph objects. This is the equivalent to digraph_union() but for a PyGraph instead of for a PyDiGraph. A new unified function union() was also added that supports both PyDiGraph and PyGraph. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

first = rustworkx.generators.path_graph(3, weights=["a_0", "node", "a_1"])
second = rustworkx.generators.cycle_graph(3, weights=["node", "b_0", "b_1"])
graph = rustworkx.graph_union(first, second, merge_nodes=True)
mpl_draw(graph)

The kwargs merge_nodes and merge_edges of digraph_union() are now optional and by default are set False.

Add a new find_node_by_weight() that finds the index of a node given a specific weight.

Bug Fixes#

Fixes the output of adj() to include neighbors that have an edge between them and the specified node, in either direction. Previously, only outbound nodes were included.

Previously, digraph_union() would incorrectly keep or delete edges if argument merge_edges is set to true. This has been fixed and an edge from the second graph will be skipped if both its endpoints were merged to nodes from the first graph and these nodes already share an edge with equal weight data. Fixed #432

Fixes the output of is_subgraph_isomorphic() if the input graphs have parallel edges. Fixed #429

Fixes the output of is_isomorphic() if the input graphs have parallel edges and an edge_matcher is specified. Fixed #429

Reduces the memory requirements of VF2 graph isomorphism algorithm by making use of a hash map with number of entries scaling as the number of edges in the graph rather than building the full adjacency matrix. Fixed #367

Previously, vf2_mapping() function when comparing 2 empty graph objects would incorrectly return an empty iterator. This has been fixed and now returns an iterator over the single valid mapping, i.e the empty mapping.

0.10.1#

Prelude#

This is a bugfix release that fixes a regression introduced in the previous 0.10.0 release. In 0.10.0 the is_isomorphic() function when comparing 2 empty graph objects would incorrectly return False.

0.10.0#

Prelude#

This release includes many new features and bug fixes. The highlights of this release are a slew of new algorithm functions including steiner_tree() and betweenness_centrality(), improvements to the isomorphism functions including the addition of vf2_mapping() to get an isomorphic mapping between two graphs, and improvements to working with graph objects such as rustworkx.PyDiGraph.substitute_node_with_subgraph() for replacing a node with a subgraph.

New Features#

Added a new algorithm function, rustworkx.unweighted_average_shortest_path_length() for computing the average shortest path length of a PyDiGraph or PyGraph object. The average shortest path length is defined as

\[a =\sum_{s,t \in V} \frac{d(s, t)}{n(n-1)}\]

where \(V\) is the set of nodes in the graph, \(d(s, t)\) is the shortest path length from \(s\) to \(t\), and \(n\) is the number of nodes in graph.

Added a new function, betweenness_centrality() to compute betweenness centrality of all nodes in a PyGraph or PyDiGraph object. The algorithm used in this function is based on:

Ulrik Brandes, “A Faster Algorithm for Betweenness Centrality”. Journal of Mathematical Sociology 25(2):163-177, 2001. DOI: 10.1080/0022250X.2001.9990249

The betweenness centrality of a node \(v\) is the sum of the fraction of all-pairs shortest paths that pass through \(v\)

\[c_B(v) =\sum_{s,t \in V} \frac{\sigma(s, t|v)}{\sigma(s, t)}\]

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 \(v \in {s, t}\), \(\sigma(s, t|v) = 0\)

For example, computing the betweenness centrality for all nodes in a 5x5 grid graph and using that to color the nodes in a graph visualization:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.generators.grid_graph(5, 5) btw = rustworkx.betweenness_centrality(graph) # Color nodes in graph visualization with betweenness centrality colors = [] for i in range(len(graph)): colors.append(btw[i]) mpl_draw(graph, node_color=colors)

Added two new algorithm functions, dag_weighted_longest_path_length() and dag_weighted_longest_path(), to find the longest path and the length of the longest path in a PyDiGraph object without any cycles. These new functions are basically equivalent to dag_longest_path() and dag_longest_path_length() except for two key differences. First the weight_fn parameter is required for dag_weighted_longest_path_length() and dag_weighted_longest_path_length() while it is optional in dag_longest_path() and dag_longest_path_length(). Secondly, dag_weighted_longest_path() and dag_weighted_longest_path_length() work with float weights (dag_weighted_longest_path_length() returns a float and the weight_fn callback for both is expected to return a float) while dag_longest_path() and dag_longest_path_length() works with an unsigned int.

Added a new method, edge_subgraph(), to the PyDiGraph and PyGraph (edge_subgraph()) to get an edge induced subgraph from a given graph object.

Added new generator functions, rustworkx.generators.heavy_hex_graph() and rustworkx.generators.directed_heavy_hex_graph(), for constructing a heavy hex graphs from: https://journals.aps.org/prx/abstract/10.1103/PhysRevX.10.011022 For example:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.generators.directed_heavy_hex_graph(3) mpl_draw(graph)

Added new generator functions, rustworkx.generators.heavy_square_graph() and rustworkx.generators.directed_heavy_square_graph(), for generating heavy square graphs from: https://journals.aps.org/prx/abstract/10.1103/PhysRevX.10.011022 For example:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.generators.directed_heavy_square_graph(3) mpl_draw(graph)

A new kwarg, labels, has been added to the rustworkx.PyDiGraph.read_edge_list() and rustworkx.PyGraph.read_edge_list() constructor methods. When this kwarg is set to True the first 2 elements on a line in an edge list file are treated as string labels uniquely identifying a node instead of integer node indices. For example:

import tempfile

import rustworkx
from rustworkx.visualization import mpl_draw

with tempfile.NamedTemporaryFile("wt") as fd:
  fd.write('a b first_edge\n')
  fd.write('b c second_edge\n')
  fd.flush()
  graph = rustworkx.PyDiGraph.read_edge_list(fd.name, labels=True)

mpl_draw(graph, with_labels=True, labels=str, edge_labels=str)

The adjacency_matrix() function has a new kwarg null_value which is used to adjust the value used in the output matrix representing the absence of an edge. This can be set to any float value and if not specified the default value of 0.0 is still used. For example:

import numpy as np
import rustworkx

graph = rustworkx.generators.cycle_graph(4)
distance_matrix = rustworkx.adjacency_matrix(graph, null_value=np.inf)
print(distance_matrix)

[[inf  1. inf  1.]
 [ 1. inf  1. inf]
 [inf  1. inf  1.]
 [ 1. inf  1. inf]]

The distance_matrix() function has a new kwarg null_value which is used to adjust the value used in the output matrix representing the absence of a path. This can be set to any float value and if not specified the default value of 0.0 is still used. For example:

import numpy as np
import rustworkx

graph = rustworkx.generators.cycle_graph(4)
graph.add_node(None)
graph.add_node(None)
distance_matrix = rustworkx.distance_matrix(graph, null_value=np.inf)
print(distance_matrix)

[[ 0.  1.  2.  1. inf inf]
 [ 1.  0.  1.  2. inf inf]
 [ 2.  1.  0.  1. inf inf]
 [ 1.  2.  1.  0. inf inf]
 [inf inf inf inf  0. inf]
 [inf inf inf inf inf  0.]]

A new method, substitute_node_with_subgraph(), to the PyDiGraph class. This method is used to replace a node in a PyDiGraph object with another PyDiGraph object. For example, first creating a new graph:

import rustworkx
from rustworkx.visualization import mpl_draw

original_graph = rustworkx.generators.directed_path_graph(5)
mpl_draw(original_graph, with_labels=True)

then create another graph to use in place of a node:

other_graph = rustworkx.generators.directed_star_graph(25)
mpl_draw(other_graph)

finally replace a node in the original graph with the second graph:

def edge_map_fn(_source, _target, _weight):
    return 0

node_mapping = original_graph.substitute_node_with_subgraph(2, other_graph, edge_map_fn)
print(node_mapping)
mpl_draw(original_graph, with_labels=True)

NodeMap{0: 5, 1: 6, 2: 7, 3: 8, 4: 9, 5: 10, 6: 11, 7: 12, 8: 13, 9: 14, 10: 15, 11: 16, 12: 17, 13: 18, 14: 19, 15: 20, 16: 21, 17: 22, 18: 23, 19: 24, 20: 25, 21: 26, 22: 27, 23: 28, 24: 29}

Added a new method, to_directed(), to the PyGraph class. This method is used to generate a new rustworkx.PyDiGraph directed graph object from an undirected PyGraph object. The output rustworkx.PyDiGraph object will have a bidirectional edge pair for each edge in the original PyGraph object.

Added a new kwarg, null_value to the rustworkx.PyDiGraph.from_adjacency_matrix() and rustworkx.PyGraph.from_adjacency_matrix() which is used to optionally change the null value in the matrix treated as the absence of an edge. By default 0.0 is used. For example:

import numpy as np
import rustworkx
from rustworkx.visualization import mpl_draw

matrix = np.array([[np.nan, 1, 1], [1, np.nan, 1], [1, 1, 0]], dtype=np.float64)
graph = rustworkx.PyDiGraph.from_adjacency_matrix(matrix, null_value=np.nan)
mpl_draw(graph, with_labels=True, edge_labels=str)

The floyd_warshall() function has been completely reworked and now works with a PyGraph object in addition to the PyDiGraph objects already supported. It has a new kwarg weight_fn which is used to support custom edge weighting, and also now executes in parallel.

Added two new functions, graph_floyd_warshall() and digraph_floyd_warshall(), which are typed versions of the existing floyd_warshall().

Two new kwargs, multigraph and weight_combo_fn, were added to rustworkx.PyDiGraph.to_undirected() method. These two options can be used to condense parallel edges into a single edge in the output graph. By setting multigraph to True any parallel edges will be condensed into a single edge in the output PyGraph. The weight_combo_fn kwarg is used to control how the parallel edges are condensed. If it’s not specified the weight/data payload of the edge with the largest edge index will be used. If weight_combo_fn is specified it will be passed the weight objects for any parallel edges in the graph and the returned object will be used as the weight for the condensed edge. For example:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.PyDiGraph() graph.extend_from_weighted_edge_list([ (0, 1, 25), (1, 0, 10), (1, 2, 10), (2, 1, 100), (2, 3, 50), (3, 2, 40), ]) undirected_graph = graph.to_undirected( multigraph=False, weight_combo_fn=max ) mpl_draw(undirected_graph, with_labels=True, edge_labels=str)

A new kwarg, induced, was added to is_subgraph_isomorphic(). If set to True the function will check if the second graph is isomorphic to a node-induced subgraph of the first. If set to False it will check if the second graph is isomorphic to a subgraph of the first. By default this is set to True.

Added a new function, num_shortest_paths_unweighted() (and it’s per type variants digraph_num_shortest_paths_unweighted() and graph_num_shortest_paths_unweighted()), which is used to get a count of unweighted shortest paths from a given node to all other nodes (where path exists) in the graph.

Added a new method, has_parallel_edges(), to the PyDiGraph and PyGraph (has_parallel_edges()) classes. This method returns True if there are parallel edges in the graph and False otherwise.

Added a new function, metric_closure(), which is used to generate the metric closure of a given graph. This function is used in the computation for calculating the Steiner Tree using the algorithm from Kou, Markowsky & Berman (1981). “A fast algorithm for Steiner trees”. https://link.springer.com/article/10.1007/BF00288961

Added a new function, steiner_tree(), which is used to generate an approximation of the minimum Steiner tree using the algorithm from: Kou, Markowsky & Berman (1981). “A fast algorithm for Steiner trees”. https://link.springer.com/article/10.1007/BF00288961

Added a new function, rustworkx.vf2_mapping(), which will use the vf2 isomorphism algorithm (which is also used for rustworkx.is_isomorphic() and rustworkx.is_subgraph_isomorphic()) to return an iterator over all valid isomorphic mappings between two graphs. For example:

import rustworkx

graph = rustworkx.generators.directed_grid_graph(10, 10)
other_graph = rustworkx.generators.directed_grid_graph(4, 4)
vf2 = rustworkx.vf2_mapping(graph, other_graph, subgraph=True)
try:
  mapping = next(vf2)
  print(mapping)
except StopIteration:
  pass

NodeMap{0: 0, 1: 1, 2: 2, 3: 3, 10: 4, 11: 5, 12: 6, 13: 7, 20: 8, 21: 9, 22: 10, 23: 11, 30: 12, 31: 13, 32: 14, 33: 15}

Added a new kwarg, call_limit to rustworkx.is_isomorphic() and rustworkx.is_subgraph_isomorphic() which is used to set an upper bound on the number of states that VF2 algorithm visits while searching for a solution. If it exceeds this limit, the algorithm will stop and return false.

Added a new kwarg, weight_fn, to the rustworkx.dag_longest_path() and rustworkx.dag_longest_path_length() functions. These kwargs are used to set a weight function callable that will be called as the function traverses the graph to dynamically adjust the weight of an edge. For example:

import retwork as rx

dag = rx.PyDiGraph()
dag.extend_from_weighted_edge_list(
      [
          (0, 1, 1),
          (0, 3, 1),
          (3, 4, 1),
          (4, 5, 1),
          (1, 2, 1),
          (0, 1, 3),
      ]
  )
  longest_path = rx.dag_longest_path(dag, lambda _, __, weight: weight)
  path_length = rx.dag_longest_path_length(dag, lambda_, __, weight: weight)
  assert [0, 1, 2] == longest_path
  assert 4 == weight

Upgrade Notes#

The floyd_warshall() function no longer returns a dict and instead will return a rustworkx.AllPairsPathLengthMapping object. This new return type is much faster to build and it implements the Python mapping protocol in a read-only fashion. This change should mostly be compatible unless explicit type checking or mutating the output was done.

Bug Fixes#

Support for negative weights in the rustworkx.PyDiGraph.from_adjacency_matrix() and rustworkx.PyGraph.from_adjacency_matrix() methods has been fixed. Previously, if a negative weight were used it would be incorrectly treated as a null value and no edge was added to the graph. This has been corrected so that a negative value in the input matrix is now treated as an edge with a negative weight. For example:
import numpy as np import rustworkx from rustworkx.visualization import mpl_draw matrix = np.array([[0, -1, -1], [1, 0, -1], [1, 1, 0]], dtype=np.float64) graph = rustworkx.PyDiGraph.from_adjacency_matrix(matrix) mpl_draw(graph, with_labels=True, edge_labels=str)
Fixed #408

Fixed an issue with the Dijkstra path functions:
where the returned paths could be incorrect in cases where there were multiple paths between nodes and an edge weight callback function was specified. Fixed #387

Fixes the output of number_weakly_connected_components() in case of holes in the graph node indices.

Previously, the graphs created by rustworkx.generators.directed_hexagonal_lattice_graph() and rustworkx.generators.hexagonal_lattice_graph() contained holes in the node indices. This has been fixed by generating graphs with consecutive node indexes. Fixed #373

The iteration order of the output dict returned by graph_greedy_color() was previously not deterministic and multiple executions with the same input would return identical dictionaries that had different insertion orders which would result in different iteration order. This has been fixed so that the iteration order of the output dictionary is now deterministic. Fixed #347

0.9.0#

Prelude#

This release is a new feature release that includes a plethora of new features and bug fixes. The highlights of this release are the introduction of the retwork.visualization module, which includes a matplotlib based drawer (mpl_draw()) and a graphviz based drawer (graphviz_draw()), and layout functions such as spring_layout() for generating a layout in visualization. Additionally, the generator functions in rustworkx.generators have been expanded to include new graph generators, and many new algorithm functions have been added.

New Features#

Added new generator functions, rustworkx.generators.binomial_tree_graph() and rustworkx.generators.directed_binomial_tree_graph(), for constructing a binomial tree graph. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.binomial_tree_graph(4)
mpl_draw(graph)

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.directed_binomial_tree_graph(4)
mpl_draw(graph)

A new method, copy(), has been added to the PyDiGraph and PyGraph (copy()) classes. This method will return a shallow copy of a graph object.

Add a new function, core_number() for computing the core number of each node in a PyGraph or a PyDiGraph, where for directed graphs the degrees are calculated as in_degree + out_degree.

Added a new method, edge_indices(), to the PyDiGraph and PyGraph (edge_indices()) that will return a list of edge indices for every edge in the graph object.

Added a new custom return type EdgeIndices which is returned by rustworkx.PyDiGraph.edge_indices and rustworkx.PyGraph.edge_indices. It is equivalent to a read-only list of integers that represent a list of edge indices.

Added a new method, num_nodes(), to the PyDiGraph and PyGraph (num_nodes()) for returning the number of nodes in the graph.

Added a new method, num_edges(), to the PyDiGraph and PyGraph (num_edges()) for returning the number of edges in the graph.

A new function rustworkx.networkx_converter() has been added. This function takes in a networkx Graph object and will generate an equivalent PyGraph or PyDiGraph object. While this function is provided as a convenience for users of both rustworkx and networkx, networkx will not be added as a dependency of rustworkx (which precludes a rustworkx->networkx converter, see Converting from a networkx graph for example code on how to build this yourself).

Added new function random_geometric_graph() which can be used to generate random geometric graphs. For example:

import rustworkx
from rustworkx.visualization import mpl_draw


graph = rustworkx.random_geometric_graph(8, .95, 5)
mpl_draw(graph)

Added a new layout function, rustworkx.random_layout() (and it’s equivalent per type variants rustworkx.graph_random_layout() and rustworkx.diraph_random_layout()) to generate a random layout which can be used for visualizations. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.directed_grid_graph(5, 5)
layout = rustworkx.random_layout(graph)
mpl_draw(graph, pos=layout)

or with the graphviz_draw() function:

import rustworkx
from rustworkx.visualization import graphviz_draw

graph = rustworkx.generators.directed_grid_graph(5, 5)
layout = rustworkx.random_layout(graph)
for i in range(25):
  graph[i] = i

def node_attr_fn(node):
    point = layout[node]
    return {
        "shape": "circle",
        "pos": '"%s,%s!"' % (point[0], point[1]),
        "fillcolor": "yellow",
        "style": "filled",
        "fixedsize": "true"
    }

graphviz_draw(graph, node_attr_fn=node_attr_fn, method='fdp')

A new custom return class, rustworkx.Pos2DMapping, has been added. This class will be returned by layout functions and is a drop in replacement for an immutable read-only dictionary of the form:
```
{1: [0.1, 0.5]}
```
where the keys are node indices and the values are a 2 element sequence that represents the position for the node.

Added a new function, is_subgraph_isomorphic(), to determine if two graphs of the same type (either PyGraph or PyDiGraph) are induced subgraph isomorphic.

A new function, transitivity(), was added to calculate the transitivity coefficient of a PyGraph or a PyDiGraph object.

Added a new function complement() to calculate the graph complement of a PyGraph or PyDiGraph.

Two new return types PathMapping and PathLengthMapping. These classes are returned from functions that previously returned a dictionary of paths or a dictionary of path lengths. They both implement the Python mapping protocol and can be used inplace as a read-only dict.

New functions have been added to compute the shortest path and shortest path lengths for all nodes in a graph object:

Added a new method, edge_index_map(), to the PyDiGraph and PyGraph (edge_index_map()) that will return a read-only mapping of edge indices to a tuple of the form (source_node_index, target_node_index, weight/data payload) for every edge in the graph object.

Added a new custom return type EdgeIndexMap which is returned by rustworkx.PyDiGraph.edge_index_map() and rustworkx.PyGraph.edge_index_map(). It is equivalent to a read-only dict/mapping that represent a mapping of edge indices to the edge tuple.

The is_isomorphic() function has been expanded so it can now also take in a PyGraph in addition to the the PyDiGraph already supported.

The is_isomorphic() function now has two new optional kwargs node_matcher and edge_matcher which can be used to specify functions to use for comparing node and edge data payloads.

The is_isomorphic_node_match() function has been expanded so it can take in a PyGraph in addition to the PyDiGraph it already supported.

Added new generator functions, rustworkx.generators.directed_hexagonal_lattice_graph() and rustworkx.generators.hexagonal_lattice_graph(), for constructing a hexagonal lattice graph. For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.directed_hexagonal_lattice_graph(3, 3)
mpl_draw(graph)

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.hexagonal_lattice_graph(3, 3)
mpl_draw(graph)

Added two new methods, find_successors_by_edge() and find_predecessors_by_edge(), to PyDiGraph. These methods efficiently retrieve the neighbors in the graph which are connected to a node with edges matching a filter function.

The is_isomorphic() and is_isomorphic_node_match() functions have a new kwarg, id_order which is used to adjust the node matching order used. If you set id_order=False then the matching order used is the heuristic matching order proposed in the VF2++ paper. If you want to retain use the order based on node ids, you can set id_order=True which is the default behavior.

Added new function, minimum_spanning_tree(), to calculate the minimum spanning tree of a PyGraph object and return the MST as a new PyGraph object.

Added a new function, minimum_spanning_edges(), to calculate the minimum spanning tree of a PyGraph object and return the WeightedEdgeList for the weighted edge list of the MST of a graph.

A new module, rustworkx.visualization has been added. This module will contain various functions used for visualizing rustworkx graphs.

A new visualization function, rustworkx.visualization.mpl_drawer(), for visualizing graphs with Matplotlib has been added. This function requires that matplotlib, which is not a dependency of rustworkx, to be installed. To install matplotlib you can either use pip install matplotlib or when you install rustworkx pip install 'rustworkx[mpl]'. This function can take any rustworkx graph object, a PyGraph or PyDiGraph and visualize them with various options to tweak the output. For example, a basic graph without any labels is:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.generators.grid_graph(4, 6) mpl_draw(graph)
or to change the colors:
import rustworkx from rustworkx.visualization import mpl_draw graph = rustworkx.generators.grid_graph(4, 6) mpl_draw(graph, node_color='r', edge_color='#00FFFF')
Refer to the function documentation for a full list of options.

A new kwarg, multigraph was added to the PyDiGraph generator functions:
This can be used to set whether the generated PyDiGraph objects are multi graphs or not (to set whether parallel edges can be added or not). By default this is set to True.

Added a new Graphviz based drawer function, graphviz_draw(), to the rustworkx.visualization module. This function requires that Graphviz is installed locally and adds two new optional dependencies, pydot which is used to call Graphviz and Pillow to interact with the generated image files. The optional dependencies can be installed either with pip install pydot pillow` or when installing rustworkx with ``pip install 'rustworkx[graphviz]'. This function wraps the to_dot() method to generate a dot representation of the graph and will call Graphviz to generate a visualization of the graph. For example:
import rustworkx from rustworkx.visualization import graphviz_draw def node_attr(node): if node == 0: return {'color': 'yellow', 'fillcolor': 'yellow', 'style': 'filled'} if node % 2: return {'color': 'blue', 'fillcolor': 'blue', 'style': 'filled'} else: return {'color': 'red', 'fillcolor': 'red', 'style': 'filled'} graph = rustworkx.generators.directed_star_graph(weights=list(range(32))) graphviz_draw(graph, node_attr_fn=node_attr, method='sfdp')

Four simple layout functions were added:

bipartite_layout()

circular_layout()

shell_layout()

spiral_layout()

These can be used to adjust the layout used in visualizations, for example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.path_graph(weights=list(range(24)))
layout = rustworkx.bipartite_layout(graph, set(range(12)))
mpl_draw(graph, pos=layout)

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.path_graph(weights=list(range(24)))
layout = rustworkx.circular_layout(graph)
mpl_draw(graph, pos=layout)

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.star_graph(25)
layout = rustworkx.shell_layout(graph)
mpl_draw(graph, pos=layout)

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.generators.path_graph(weights=list(range(24)))
layout = rustworkx.spiral_layout(graph)
mpl_draw(graph, pos=layout)

Or with the graphviz_drawer() function:

import rustworkx
from rustworkx.visualization import graphviz_draw

graph = rustworkx.generators.path_graph(weights=list(range(24)))
layout = rustworkx.spiral_layout(graph)

def node_attr_fn(node):
    point = layout[node]
    return {
        "shape": "circle",
        "pos": '"%s,%s!"' % (point[0], point[1]),
        "fillcolor": "yellow",
        "style": "filled",
    }

graphviz_draw(graph, node_attr_fn=node_attr_fn, method='fdp')

Added a new function, spring_layout() to generate layouts for PyGraph and PyDiGraph using the Fruchterman-Reingold force-directed algorithm. This layout method is used by default for the mpl_drawer() visualization function. You can also explicitly use it when calling mpl_drawer() and graphviz_drawer(). For example:

import rustworkx
from rustworkx.visualization import mpl_draw

graph = rustworkx.random_geometric_graph(15, 1.42)
layout = rustworkx.spring_layout(graph, adaptive_cooling=False)
mpl_draw(graph, pos=layout)

and with the graphviz drawer:

import rustworkx
from rustworkx.visualization import graphviz_draw

graph = rustworkx.random_geometric_graph(15, 1.42)
layout = rustworkx.spring_layout(graph, adaptive_cooling=False)
for i in range(15):
  graph[i] = i

def node_attr_fn(node):
    point = layout[node]
    return {
        "shape": "circle",
        "pos": '"%s,%s!"' % (point[0], point[1]),
        "fillcolor": "yellow",
        "style": "filled",
        "fixedsize": "true"
    }

graphviz_draw(graph, node_attr_fn=node_attr_fn, method='fdp')

A new method, write_edge_list(), has been added to the PyDiGraph and PyGraph (write_edge_list()) classes. This method is used to write an edge list file representing the graph object. The output format is such that it can be used as an input to rustworkx.PyDiGraph.read_edge_list() and rustworkx.PyGraph.read_edge_list().

Upgrade Notes#

The functions:
no longer are returning a dict and instead are returning rustworkx.PathLengthMapping objects. This new return type is much faster to build and it implements the python mapping protocol in a read-only fashion and should not be noticeable unless explicit type checking or mutating the result were done.

The functions:
no longer are returning a dict and instead are returning rustworkx.PathMapping objects. This new return type is much faster to build and it implements the python mapping protocol in a read-only fashion and should not be noticeable unless explicit type checking or mutating the result were done.

Bug Fixes#

Fixed an issue where calling rustworkx.PyDiGraph.successor_indices() or rustworkx.PyDiGraph.predecessor_indices() would raise a RuntimeError exception if they were called in a context where rustworkx is already working with a reference to a PyDiGraph (primarily if it were called in a callback function for another PyDiGraph method).

Fix bug in floyd_warshall_numpy() in which the dispatcher mistakenly called adjacency_matrix() instead of graph_floyd_warshall_numpy() and digraph_floyd_warshall_numpy().

0.8.0#

Prelude#

This release includes several new features and bug fixes. The main features for this release are some usability improvements including the introduction of new methods for interacting with edges, constructing graphs from adjacency matrices, and universal functions that are not strictly typed and will work with either a PyGraph or PyDiGraph object. It also includes new algorithm functions around matchings for a PyGraph, including a function to find the maximum weight matching. This is also the first release to include support and publishing of precompiled binaries for Apple Arm CPUs on MacOS.

New Features#

A new constructor method from_adjacency_matrix() has been added to the PyDiGraph and PyGraph (from_adjacency_matrix()) classes. It enables creating a new graph from an input adjacency_matrix. For example:

import os
import tempfile

import numpy as np
import pydot
from PIL import Image

import rustworkx


# Adjacency matrix for directed outward star graph:
adjacency_matrix = np.array([
    [0., 1., 1., 1., 1.],
    [0., 0., 0., 0., 0.],
    [0., 0., 0., 0., 0.],
    [0., 0., 0., 0., 0.],
    [0., 0., 0., 0., 0.]])
# Create a graph from the adjacency_matrix:
graph = rustworkx.PyDiGraph.from_adjacency_matrix(adjacency_matrix)
# Draw graph
dot_str = graph.to_dot(
    lambda node: dict(
    color='black', fillcolor='lightblue', style='filled'))
dot = pydot.graph_from_dot_data(dot_str)[0]
with tempfile.TemporaryDirectory() as tmpdirname:
    tmp_path = os.path.join(tmpdirname, 'dag.png')
    dot.write_png(tmp_path)
    image = Image.open(tmp_path)
    os.remove(tmp_path)
image

A new algorithm function, is_matching(), was added to check if a matching set is valid for given PyGraph object.

A new algorithm function, is_maxmimal_matching(), was added to check if a matching set is valid and maximal for a given PyGraph object.

Add a new function, max_weight_matching() for computing the maximum-weighted matching for a PyGraph object.

The PyGraph and PyDiGraph constructors now have a new kwarg multigraph which can optionally be set to False to disallow adding parallel edges to the graph. With multigraph=False if an edge is attempted to be added where one already exists it will update the weight for the edge with the new value. For example:

import os
import tempfile

import pydot
from PIL import Image

import rustworkx as rx

graph = rx.PyGraph(multigraph=False)
graph.extend_from_weighted_edge_list([(0, 1, -1), (1, 2, 0), (2, 0, 1)])
dot = pydot.graph_from_dot_data(
    graph.to_dot(edge_attr=lambda e:{'label': str(e)}))[0]

with tempfile.TemporaryDirectory() as tmpdirname:
    tmp_path = os.path.join(tmpdirname, 'dag.png')
    dot.write_png(tmp_path)
    image = Image.open(tmp_path)
    os.remove(tmp_path)
image

Then trying to add an edge between 0 and 1 again will update its weight/payload.

graph.add_edge(0, 1, 42)
dot = pydot.graph_from_dot_data(
    graph.to_dot(edge_attr=lambda e:{'label': str(e)}))[0]

with tempfile.TemporaryDirectory() as tmpdirname:
    tmp_path = os.path.join(tmpdirname, 'dag.png')
    dot.write_png(tmp_path)
    image = Image.open(tmp_path)
    os.remove(tmp_path)
image

You can query whether a PyGraph allows multigraphs with the boolean attribute multigraph. The attribute can not be set outside of the constructor.

The rustworkx.generators module’s functions cycle_graph(), path_graph(), star_graph(), mesh_graph(), and grid_graph() now have a new kwarg multigraph which takes a boolean and defaults to True. When it is set to False the generated PyGraph object will have the multigraph attribute set to False meaning it will disallow adding parallel edges.

New universal functions that can take in a PyGraph or PyDiGraph instead of being class specific have been to the rustworkx API. These new functions are:

Starting with this release wheels will be published for macOS arm64. Only Python 3.9 is supported at first, because it is the only version of Python with native support for arm64 macOS.

The custom return types BFSSuccessors, NodeIndices, EdgeList, and WeightedEdgeList now implement __str__ so that running str() (for example when calling print() on the object) it will return a human readable string with the contents of the custom return type.

The custom return types BFSSuccessors, NodeIndices, EdgeList, and WeightedEdgeList now implement __hash__ so that running hash() (for example when insert them into a dict) will return a valid hash for the object. The only exception to this is for BFSSuccessors and WeightedEdgeList if they contain a Python object that is not hashable, in those cases calling hash() will raise a TypeError, just like as you called hash() on the inner unhashable object.

Two new methods, update_edge() and update_edge_by_index() were added to the rustworkx.PyDiGraph and rustworkx.PyGraph ( update_edge() and update_edge_by_index()) classes. These methods are used to update the data payload/weight of an edge in the graph either by the nodes of an edge or by edge index.

Upgrade Notes#

The minimum supported Rust version has increased to 1.41.1, you will need rustc >=1.41.1 now to build rustworkx. The previous minimum supported Rust version 1.39.0 will no longer be able to compile rustworkx. This is due to a changes in in the new versions of the pyo3 and rust-numpy libraries.

Bug Fixes#

In previous releases the Python garbage collector did not know how to interact with PyDiGraph or PyGraph objects and as a result they may never have been freed until Python exited. To fix this issue, the PyDiGraph and PyGraph classes now are integrated with Python’s garbage collector so they’ll properly be cleared when there are no more references to a graph object.

The output from rustworkx.PyDiGraph.neighbors() and rustworkx.PyGraph.neighbors() methods will no longer include duplicate entries in case of parallel edges between nodes. See #250 for more details.

In previous releases the Python garbage collector did not know how to interact with the custom return types BFSSuccessors, NodeIndices, EdgeList, and WeightedEdgeList and as a result they may never have been freed until Python exited. To fix this issue the custom return type classes now are integrated with Python’s garbage collector so they’ll properly be cleared when there are no more Python references to an object.

0.7.1#

This release includes a fix for an oversight in the previous 0.7.0 and 0.6.0 releases. Those releases both added custom return types BFSSuccessors, NodeIndices, EdgeList, and WeightedEdgeList that implemented the Python sequence protocol which were used in place of lists for certain functions and methods. However, none of those classes had support for being pickled, which was causing compatibility issues for users that were using the return in a context where it would be pickled (for example as an argument to or return of a function called with multiprocessing). This release has a single change over 0.7.0 which is to add the missing support for pickling BFSSuccessors, NodeIndices, EdgeList, and WeightedEdgeList which fixes that issue.

0.7.0#

This release includes several new features and bug fixes.

This release also dropped support for Python 3.5. If you want to use rustworkx with Python 3.5 that last version which supports Python 3.5 is 0.6.0.

New Features#

New generator functions for two new generator types, mesh and grid were added to rustworkx.generators for generating all to all and grid graphs respectively. These functions are: mesh_graph(), directed_mesh_graph(), grid_graph(), and directed_grid_graph()
A new function, rustworkx.digraph_union(), for taking the union between two PyDiGraph objects has been added.
A new PyDiGraph method merge_nodes() has been added. This method can be used to merge 2 nodes in a graph if they have the same weight/data payload.
A new PyDiGraph method find_node_by_weight() which can be used to lookup a node index by a given weight/data payload.
A new return type NodeIndices has been added. This class is returned by functions and methods that return a list of node indices. It implements the Python sequence protocol and can be used as list.
Two new return types EdgeList and WeightedEdgeList. These classes are returned from functions and methods that return a list of edge tuples and a list of edge tuples with weights. They both implement the Python sequence protocol and can be used as a list
A new function collect_runs() has been added. This function is used to find linear paths of nodes that match a given condition.

Upgrade Notes#

Support for running rustworkx on Python 3.5 has been dropped. The last release with support for Python 3.5 is 0.6.0.
The rustworkx.PyDiGraph.node_indexes(), rustworkx.PyDiGraph.neighbors(), rustworkx.PyDiGraph.successor_indices(), rustworkx.PyDiGraph.predecessor_indices(), rustworkx.PyDiGraph.add_nodes_from(), rustworkx.PyGraph.node_indexes(), rustworkx.PyGraph.add_nodes_from(), and rustworkx.PyGraph.neighbors() methods and the dag_longest_path(), topological_sort(), graph_astar_shortest_path(), and digraph_astar_shortest_path() functions now return a NodeIndices object instead of a list of integers. This should not require any changes unless explicit type checking for a list was used.
The rustworkx.PyDiGraph.edge_list(), and rustworkx.PyGraph.edge_list() methods and digraph_dfs_edges(), graph_dfs_edges(), and digraph_find_cycle() functions now return an EdgeList object instead of a list of integers. This should not require any changes unless explicit type checking for a list was used.
The rustworkx.PyDiGraph.weighted_edge_list(), rustworkx.PyDiGraph.in_edges(), rustworkx.PyDiGraph.out_edges(), and rustworkx.PyGraph.weighted_edge_list methods now return a WeightedEdgeList object instead of a list of integers. This should not require any changes unless explicit type checking for a list was used.

Fixes#

BFSSuccessors objects now can be compared with == and != to any other Python sequence type.
The built and published sdist packages for rustworkx were previously not including the Cargo.lock file. This meant that the reproducible build versions of the rust dependencies were not passed through to source. This has been fixed so building from sdist will always use known working versions that we use for testing in CI.

0.6.0#

This release includes a number of new features and bug fixes. The main focus of this release was to expand the rustworkx API functionality to include some commonly needed functions that were missing.

This release is also the first release to provide full support for running with Python 3.9. On previous releases Python 3.9 would likely work, but it would require building rustworkx from source. Also this will likely be the final release that supports Python 3.5.

New Features#

Two new functions, digraph_k_shortest_path_lengths() and graph_k_shortest_path_lengths(), for finding the k shortest path lengths from a node in a PyDiGraph and PyGraph.
A new method, is_symmetric(), to the PyDiGraph class. This method will check whether the graph is symmetric or not
A new kwarg, as_undirected, was added to the digraph_floyd_warshall_numpy() function. This can be used to treat the input PyDiGraph object as if it was undirected for the generated output matrix.
A new function, digraph_find_cycle(), which will return the first cycle during a depth first search of a PyDiGraph object.
Two new functions, directed_gnm_random_graph() and undirected_gnm_random_graph(), for generating random \(G(n, m)\) graphs.
A new method, remove_edges_from(), was added to PyDiGraph and PyGraph (removed_edges_from()). This can be used to remove multiple edges from a graph object in a single call.
A new method, subgraph(), was added to PyDiGraph and PyGraph (subgraph()) which takes in a list of node indices and will return a new object of the same type representing a subgraph containing the node indices in that list.
Support for running with Python 3.9
A new method, to_undirected(), was added to PyDiGraph. This method will generate an undirected PyGraph object from the PyDiGraph object.
A new kwarg, bidirectional, was added to the directed generator functions directed_cycle_graph(), directed_path_graph(), and directed_star_graph(). When set to True the directed graphs generated by these functions will add edges in both directions.
Added two new functions, is_weakly_connected() and weakly_connected_components(), which will either check if a PyDiGraph object is weakly connected or return the list of the weakly connected components of an input PyDiGraph.
The weight_fn kwarg for graph_adjacency_matrix(), digraph_adjacency_matrix(), graph_floyd_warshall_numpy(), and digraph_floyd_warshall_numpy() is now optional. Previously, it always had to be specified when calling these function. But, instead you can now rely on a default weight float (which defaults to 1.0) to be used for all the edges in the graph.
Add a neighbors() method to PyGraph and PyDiGraph (neighbors()). This function will return the node indices of the neighbor nodes for a given input node.
Two new methods, successor_indices() and predecessor_indices(), were added to PyDiGraph. These methods will return the node indices for the successor and predecessor nodes of a given input node.
Two new functions, graph_distance_matrix() and digraph_distance_matrix(), were added for generating a distance matrix from an input PyGraph and PyDiGraph.
Two new functions, digraph_dijkstra_shortest_paths() and graph_dijkstra_shortest_path(), were added for returning the shortest paths from a node in a PyDiGraph and a PyGraph object.
Four new methods, insert_node_on_in_edges(), insert_node_on_out_edges(), insert_node_on_in_edges_multiple(), and insert_node_on_out_edges_multiple() were added to PyDiGraph. These functions are used to insert an existing node in between an reference node(s) and all it’s predecessors or successors.
Two new functions, graph_dfs_edges() and digraph_dfs_edges(), were added to get an edge list in depth first order from a PyGraph and PyDiGraph.

Upgrade Notes#

The numpy arrays returned by graph_floyd_warshall_numpy(), digraph_floyd_warshall_numpy(), digraph_adjacency_matrix(), and graph_adjacency_matrix() will now be in a contiguous C array memory layout. Previously, they would return arrays in a column-major fortran layout. This was change was made to make it easier to interface the arrays returned by these functions with other C Python extensions. There should be no change when interacting with the numpy arrays via numpy’s API.
The bfs_successors() method now returns an object of a custom type BFSSuccessors instead of a list. The BFSSuccessors type implements the Python sequence protocol so it can be used in place like a list (except for where explicit type checking is used). This was done to defer the type conversion between Rust and Python since doing it all at once can be a performance bottleneck especially for large graphs. The BFSSuccessors class will only do the type conversion when an element is accessed.

Fixes#

When pickling PyDiGraph objects the original node indices will be preserved across the pickle.
The random \(G(n, p)\) functions, directed_gnp_random_graph() and undirected_gnp_random_graph(), will now also handle exact 0 or 1 probabilities. Previously it would fail in these cases. Fixes #172

0.5.0#

This release include a number of new features and bug fixes. The main focus of the improvements of this release was to increase the ease of interacting with graph objects. This includes adding support for generating dot output which can be used with graphviz (or similar tools) for visualizing graphs adding more methods to query the state of graph, adding a generator module for easily creating graphs of certain shape, and implementing the mapping protocol so you can directly interact with graph objects.

New Features#

A new method, to_dot(), was added to PyGraph and PyDiGraph (to_dot()). It will generate a dot format representation of the object which can be used with Graphivz (or similar tooling) to generate visualizations of graphs.
Added a new function, strongly_connected_components(), to get the list of strongly connected components of a PyDiGraph object.
A new method, compose(), for composing another graph object of the same type into a graph was added to PyGraph and PyDiGraph (compose()).
The PyGraph and PyDigraph classes now implement the Python mapping protocol for interacting with graph nodes. You can now access and interact with node data directly by using standard map access patterns in Python. For example, accessing a graph like graph[1] will return the weight/data payload for the node at index 1.
A new module, rustworkx.generators, has been added. Functions in this module can be used for quickly generating graphs of certain shape. To start it includes:
A new method, remove_node_retain_edges(), has been added to the PyDiGraph class. This method can be used to remove a node and add edges from its predecessors to its successors.
Two new methods, edge_list() and weighted_edge_list(), for getting a list of tuples with the edge source and target (with or without edge weights) have been added to PyGraph and PyDiGraph (edge_list() and weighted_edge_list())
A new function, cycle_basis(), for getting a list of cycles which form a basis for cycles of a PyGraph object.
Two new functions, graph_floyd_warshall_numpy() and digraph_floyd_warshall_numpy(), were added for running the Floyd Warshall algorithm and returning all the shortest path lengths as a distance matrix.
A new constructor method, read_edge_list(), has been added to PyGraph and PyDigraph (read_edge_list()). This method will take in a path to an edge list file and will read that file and generate a new object from the contents.
Two new methods, extend_from_edge_list() and extend_from_weighted_edge_list() has been added to PyGraph and PyDiGraph (extend_from_edge_list() and extend_from_weighted_edge_list()). This method takes in an edge list and will add both the edges and nodes (if a node index used doesn’t exist yet) in the list to the graph.

Fixes#

The limitation with the is_isomorphic() and is_isomorphic_node_match() functions that would cause segfaults when comparing graphs with node removals has been fixed. You can now run either function with any PyDiGraph/PyDAG objects, even if there are node removals. Fixes #27
If an invalid node index was passed as part of the first_layer argument to the layers() function it would previously raise a PanicException that included a Rust backtrace and no other user actionable details which was caused by an unhandled error. This has been fixed so that an IndexError is raised and the problematic node index is included in the exception message.

0.4.0#

This release includes many new features and fixes, including improved performance and better documentation. But, the biggest change for this release is that this is the first release of rustworkx that supports compilation with a stable released version of rust. This was made possible thanks to all the hard work of the PyO3 maintainers and contributors in the PyO3 0.11.0 release.

New Features#

A new class for undirected graphs, PyGraph, was added.
2 new functions graph_adjacency_matrix() and digraph_adjacency_matrix() to get the adjacency matrix of a PyGraph and PyDiGraph object.
A new PyDiGraph method, find_adjacent_node_by_edge(), was added. This is used to locate an adjacent node given a condition based on the edge between them.
New methods, add_nodes_from(), add_edges_from(), add_edges_from_no_data(), and remove_nodes_from() were added to PyDiGraph. These methods allow for the addition (and removal) of multiple nodes or edges from a graph in a single call.
A new function, graph_greedy_color(), which is used to return a coloring map from a PyGraph object.
2 new functions, graph_astar_shortest_path() and digraph_astar_shortest_path(), to find the shortest path from a node to a specified goal using the A* search algorithm.
2 new functions, graph_all_simple_paths() and digraph_all_simple_paths(), to return a list of all the simple paths between 2 nodes in a PyGraph or a PyDiGraph object.
2 new functions, directed_gnp_random_graph() and undirected_gnp_random_graph(), to generate \(G_{np}\) random PyDiGraph and PyGraph objects.
2 new functions, graph_dijkstra_shortest_path_lengths() and digraph_dijkstra_shortest_path_lengths(), were added for find the shortest path length between nodes in PyGraph or PyDiGraph object using Dijkstra’s algorithm.

Upgrade Notes#

The PyDAG class was renamed PyDiGraph to better reflect it’s functionality. For backwards compatibility PyDAG still exists as a Python subclass of PyDiGraph. No changes should be required for existing users.
numpy is now a dependency of rustworkx. This is used for the adjacency matrix functions to return numpy arrays. The minimum version of numpy supported is 1.16.0.

Fixes#

The rustworkx exception classes are now properly exported from the rustworkx module. In prior releases it was not possible to import the exception classes (normally to catch one being raised) requiring users to catch the base Exception class. This has been fixed so a specialized rustworkx exception class can be used.