Edge demand

.github/CODE_OF_CONDUCT.md

@@ -1,25 +1,18 @@
-
 # Contributor Covenant Code of Conduct
 
 ## Our Pledge
 
 We as members, contributors, and leaders pledge to make participation in our
-community a harassment-free experience for everyone, regardless of age, body
-size, visible or invisible disability, ethnicity, sex characteristics, gender
-identity and expression, level of experience, education, socio-economic status,
-nationality, personal appearance, race, caste, color, religion, or sexual
-identity and orientation.
-
-We pledge to act and interact in ways that contribute to an open, welcoming,
-diverse, inclusive, and healthy community.
+community a harassment-free experience for everyone. We pledge to act and
+interact in ways that contribute to a healthy community.
 
 ## Our Standards
 
 Examples of behavior that contributes to a positive environment for our
 community include:
 
 * Demonstrating empathy and kindness toward other people
-* Being respectful of differing opinions, viewpoints, and experiences
+* Being respectful of differing opinions, experiences, and each other's time
 * Giving and gracefully accepting constructive feedback
 * Accepting responsibility and apologizing to those affected by our mistakes,
   and learning from the experience
@@ -37,6 +30,21 @@ Examples of unacceptable behavior include:
 * Other conduct which could reasonably be considered inappropriate in a
   professional setting
 
+## AI Policy
+
+Using LLMs to assist in writing, translation, and software development is fine,
+as long as the final output - whether that be code or text - is correct and up
+to standards.
+
+Please do not submit PRs largely based on "AI slop" code. If you cannot explain
+an implementation without going back to the LLM, do not submit it. Maintainer
+time is scarce, and we ask that you respect that by not forcing maintainers to
+be the human-reviewer-in-the-loop for a code generation tool. In any case,
+non-trivial code generation tool use must be disclosed in the PR.
+
+Issues and PRs that look like unreviewed LLM output will be closed without
+detailed feedback, and constitute a violation of this Code of Conduct.
+
 ## Enforcement Responsibilities
 
 Community leaders are responsible for clarifying and enforcing our standards of

.github/PULL_REQUEST_TEMPLATE.md

@@ -18,5 +18,6 @@ In particular:
 - Docstring additions must render correctly, including escapes and LaTeX.
 - Finally, it is essential that all contributions in this PR are license-compatible with PyVRP's MIT license.
   Please check that this PR can be included into PyVRP under the MIT license.
+- You must disclose the use of LLMs/generative AI if you used such tools to write code.
 
 </details>

.github/workflows/CD.yml

@@ -10,6 +10,8 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
+permissions: {}
+
 jobs:
   build:
     name: Build wheels on ${{ matrix.os }}
@@ -18,15 +20,15 @@ jobs:
       matrix:
         os: [ ubuntu-24.04, ubuntu-24.04-arm, windows-2022, macos-15 ]
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Build wheels
-        uses: pypa/cibuildwheel@v3.2.0
+        uses: pypa/cibuildwheel@8d2b08b68458a16aeb24b64e68a09ab1c8e82084 # v3.4.1
         with:
           package-dir: .
           output-dir: dist
-      - uses: actions/upload-artifact@v7  # upload all wheels
+      - uses: actions/upload-artifact@bbbca2ddaa5d8feaa63e36b76fdaad77386f024f # v7.0.0
         with:
           name: dist-${{ matrix.os }}
           path: ./dist/*
@@ -39,16 +41,16 @@ jobs:
     permissions:
       id-token: write  # for trusted publishing
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Install uv
-        uses: astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@cec208311dfd045dd5311c1add060b2062131d57 # v8.0.0
         with:
-          version: "0.7.5"
+          version: "0.11.3"
           enable-cache: true
           python-version: "3.13"
-      - uses: actions/download-artifact@v4  # download previously built wheels
+      - uses: actions/download-artifact@3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c # v8.0.1
         with:
           pattern: dist-*
           merge-multiple: true

.github/workflows/CI.yml

@@ -9,6 +9,8 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
+permissions: {}
+
 jobs:
   build:
     name: >
@@ -32,22 +34,22 @@ jobs:
             compiler-version: '20'
             python-version: '3.14'
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Set up Python
-        uses: actions/setup-python@v6
+        uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405 # v6.2.0
         with:
           python-version: ${{ matrix.python-version }}
       - name: Install uv
-        uses: astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@cec208311dfd045dd5311c1add060b2062131d57 # v8.0.0
         with:
-          version: "0.7.5"
+          version: "0.11.3"
           enable-cache: true
       - name: Install Python dependencies
         run: uv sync --no-install-project
       - name: Cache pre-commit
-        uses: actions/cache@v5
+        uses: actions/cache@668228422ae6a00e4ad889ee87cd7109ec5666a7 # v5.0.4
         id: cache-pre-commit
         with:
           path: ~/.cache/pre-commit/

.github/workflows/CodSpeed.yml

@@ -9,22 +9,24 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.ref }}
   cancel-in-progress: true
 
+permissions: {}
+
 jobs:
   bench:
     name: Run benchmarks
     runs-on: ubuntu-24.04
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Set up Python
-        uses: actions/setup-python@v6
+        uses: actions/setup-python@a309ff8b426b58ec0e2a45f0f869d46889d02405 # v6.2.0
         with:
           python-version: '3.13'
       - name: Install uv
-        uses: astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@cec208311dfd045dd5311c1add060b2062131d57 # v8.0.0
         with:
-          version: "0.7.5"
+          version: "0.11.3"
           enable-cache: true
       - name: Install Python dependencies
         run: uv sync --no-install-project
@@ -56,7 +58,7 @@ jobs:
       - name: Run benchmarks with CodSpeed
         if: ${{ secrets.CODSPEED_TOKEN != '' }}
         # We evaluate the microbenchmarks using the release build.
-        uses: CodSpeedHQ/action@v3
+        uses: CodSpeedHQ/action@76578c2a7ddd928664caa737f0e962e3085d4e7c # v3.8.1
         with:
           token: ${{ secrets.CODSPEED_TOKEN }}
           run: uv run pytest benchmarks/ --codspeed

.github/workflows/DOC.yml

@@ -8,20 +8,22 @@ concurrency:
   group: "pages"
   cancel-in-progress: true
 
+permissions: {}
+
 jobs:
   build:
     name: "Build documentation"
     runs-on: ubuntu-24.04
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@de0fac2e4500dabe0009e67214ff5f5447ce83dd # v6.0.2
         with:
           persist-credentials: false
       - name: Set up pages
-        uses: actions/configure-pages@v4
+        uses: actions/configure-pages@45bfe0192ca1faeb007ade9deae92b16b8254a0d # v6.0.0
       - name: Install uv
-        uses: astral-sh/setup-uv@v7
+        uses: astral-sh/setup-uv@cec208311dfd045dd5311c1add060b2062131d57 # v8.0.0
         with:
-          version: "0.7.5"
+          version: "0.11.3"
           enable-cache: true
           python-version: "3.13"
       - name: Update package index
@@ -45,7 +47,7 @@ jobs:
       - name: Build documentation
         run: uv run make html --directory=docs
       - name: Upload documentation
-        uses: actions/upload-pages-artifact@v3
+        uses: actions/upload-pages-artifact@fc324d3547104276b827a68afc52ff2a11cc49c9 # v5.0.0
         with:
           path: 'docs/build/html'
 
@@ -64,4 +66,4 @@ jobs:
     steps:
       - name: Deploy documentation
         id: deployment
-        uses: actions/deploy-pages@v4
+        uses: actions/deploy-pages@cd2ce8fcbc39b97be8ca5fce6e763baed58fa128 # v5.0.0

README.md

@@ -28,7 +28,7 @@ pip install pyvrp
 The documentation is available [here][1].
 
 > [!TIP]
-> Looking for professional support or enterprise features? [RoutingLab](https://routinglab.tech) provides consulting, custom development, and FastVRP - a production-ready route optimisation API built on PyVRP.
+> Looking for professional support or enterprise features? [RoutingLab](https://routinglab.tech) provides consulting, custom development, and PyVRP Enterprise - an extended and production-ready version of PyVRP.
 
 ### Tutorials
 

benchmarks/search/test_EdgeDemands.py

@@ -0,0 +1,124 @@
+import numpy as np
+import pytest
+
+from pyvrp import (
+    Client,
+    CostEvaluator,
+    Depot,
+    Location,
+    ProblemData,
+    RandomNumberGenerator,
+    Solution,
+    VehicleType,
+)
+from pyvrp.search import (
+    Exchange11,
+    LocalSearch,
+    PerturbationManager,
+    PerturbationParams,
+)
+
+NUM_CLIENTS = 90
+SEED = 42
+
+
+def _distance_matrix(size: int) -> np.ndarray:
+    idx = np.arange(size, dtype=np.int64)
+    mat = np.abs(idx[:, None] - idx[None, :]) + 1
+    np.fill_diagonal(mat, 0)
+    return mat
+
+
+def _edge_demand_matrix(size: int) -> np.ndarray:
+    frm = np.arange(size, dtype=np.int64)[:, None]
+    to = np.arange(size, dtype=np.int64)[None, :]
+
+    # Deterministic directed pattern with non-zero off-diagonals.
+    mat = ((frm * 17 + to * 31) % 5 == 0).astype(np.int64)
+    np.fill_diagonal(mat, 0)
+    return mat
+
+
+def _make_data(with_edge_demands: bool) -> ProblemData:
+    size = NUM_CLIENTS + 1  # includes depot
+
+    locs = [Location(idx, idx % 13) for idx in range(size)]
+    clients = [Client(client, delivery=[1]) for client in range(1, size)]
+    depots = [Depot(0)]
+
+    # Keep capacity large so edge demands do not change objective values.
+    veh_types = [VehicleType(num_available=3, capacity=[1_000_000])]
+    dist = _distance_matrix(size)
+    dur = dist.copy()
+
+    try:
+        if with_edge_demands:
+            return ProblemData(
+                locations=locs,
+                clients=clients,
+                depots=depots,
+                vehicle_types=veh_types,
+                distance_matrices=[dist],
+                duration_matrices=[dur],
+                edge_demand_matrices=[[_edge_demand_matrix(size)]],
+            )
+
+        return ProblemData(
+            locations=locs,
+            clients=clients,
+            depots=depots,
+            vehicle_types=veh_types,
+            distance_matrices=[dist],
+            duration_matrices=[dur],
+        )
+    except TypeError:
+        if with_edge_demands:
+            pytest.skip(
+                "This PyVRP build does not support edge_demand_matrices yet."
+            )
+        raise
+
+
+def _make_search(data: ProblemData):
+    rng = RandomNumberGenerator(seed=SEED)
+    neighbours = [
+        [nb for nb in range(data.num_clients) if nb != client]
+        for client in range(data.num_clients)
+    ]
+    ls = LocalSearch(
+        data,
+        rng,
+        neighbours,
+        PerturbationManager(PerturbationParams(0, 0)),
+    )
+    ls.add_operator(Exchange11(data))
+
+    # Multiple routes improve move opportunities while keeping deterministic
+    # route structure across benchmark rounds.
+    routes = [
+        list(range(0, 30)),
+        list(range(30, 60)),
+        list(range(60, data.num_clients)),
+    ]
+    sol = Solution(data, routes)
+    cost_eval = CostEvaluator([1], 0, 0)
+
+    return ls, sol, cost_eval
+
+
+@pytest.mark.parametrize(
+    "with_edge_demands",
+    [False, True],
+    ids=["without_edge_demands", "with_edge_demands"],
+)
+def test_exchange11_edge_demand_overhead(with_edge_demands, benchmark):
+    """
+    Benchmarks local-search runtime with and without edge demand matrices.
+
+    For legacy comparison, run this benchmark on ``main``:
+    ``without_edge_demands`` there is case (1), and on this branch it is case
+    (2). Case (3) is ``with_edge_demands`` on this branch.
+    """
+    data = _make_data(with_edge_demands)
+    ls, sol, cost_eval = _make_search(data)
+    benchmark(ls, sol, cost_eval, exhaustive=True)

docs/source/conf.py

@@ -172,3 +172,13 @@ def linkcode_resolve(domain: str, info: dict) -> str | None:
 
 # -- Options for EPUB output
 epub_show_urls = "footnote"
+
+
+def setup(app):
+    # Set up Rybbit analytics.
+    kwargs = {"data-site-id": "aeb758c34279"}
+    app.add_js_file(
+        "https://app.rybbit.io/api/script.js",
+        loading_method="defer",
+        **kwargs,
+    )

docs/source/dev/contributing.rst

@@ -191,6 +191,7 @@ This greatly reduces the job of maintaining and releasing the software.
 - If you are adding new functionality, you need to add it to the documentation by editing (or creating) the appropriate file in ``docs/source/``.
 - Make sure your documentation changes parse correctly.
   See the documentation in the ``docs/`` directory for details on how to build the documentation locally.
+- Disclose the use of LLMs/generative AI if you used such tools to develop the feature in your pull request.
 
 .. note::
 

docs/source/index.rst

@@ -25,7 +25,7 @@ It can be installed through *pip* via
 
 .. tip::
 
-   Looking for professional support or enterprise features? `RoutingLab <https://routinglab.tech>`_ provides consulting, custom development, and FastVRP - a production-ready route optimisation API built on PyVRP.
+   Looking for professional support or enterprise features? `RoutingLab <https://routinglab.tech>`_ provides consulting, custom development, and PyVRP Enterprise - an extended and production-ready version of PyVRP.
 
 Contents
 --------