ascent-science-aco 0.1.0__tar.gz

ascent_science_aco-0.1.0/PKG-INFO

@@ -0,0 +1,13 @@
+Metadata-Version: 2.4
+Name: ascent-science-aco
+Version: 0.1.0
+Summary: Ant colony optimizer plugin for Ascent Science SDK
+License-Expression: LicenseRef-Proprietary
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: ascent-science<0.10,>=0.9
+
+# Ascent Science ACO plugin
+
+Install with `python -m pip install ascent-science-aco`, then use
+`ascent.combinatorial.ACOOptimizer(...)`.

ascent_science_aco-0.1.0/README.md

@@ -0,0 +1,4 @@
+# Ascent Science ACO plugin
+
+Install with `python -m pip install ascent-science-aco`, then use
+`ascent.combinatorial.ACOOptimizer(...)`.

ascent_science_aco-0.1.0/pyproject.toml

@@ -0,0 +1,19 @@
+[build-system]
+requires = ["setuptools>=61"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "ascent-science-aco"
+version = "0.1.0"
+description = "Ant colony optimizer plugin for Ascent Science SDK"
+readme = "README.md"
+requires-python = ">=3.9"
+license = "LicenseRef-Proprietary"
+dependencies = ["ascent-science>=0.9,<0.10"]
+
+[project.entry-points."ascent.algorithms"]
+aco = "ascent_science_aco:register"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+include = ["ascent_science_aco*"]

ascent_science_aco-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

ascent_science_aco-0.1.0/src/ascent_science_aco/__init__.py

@@ -0,0 +1,163 @@
+"""Ant colony optimization plugin for Ascent Science."""
+
+from __future__ import annotations
+
+import math
+import random
+import time
+from dataclasses import asdict, dataclass
+
+from ascent.core import AlgorithmSpec
+from ascent.plugin_api import register_algorithm, square_matrix, tracked, validated_license
+
+__version__ = "0.1.0"
+
+
+@dataclass(frozen=True)
+class ACOResult:
+    best_route: list[int]
+    best_distance: float
+    history: list[float]
+    iterations: int
+    runtime_seconds: float
+    solver_name: str = "ACOOptimizer"
+    solver_version: str = "ascent-science-aco 0.1.0"
+    execution_location: str = "local"
+    status: str = "FEASIBLE"
+
+    @property
+    def best_solution(self):
+        return self.best_route
+
+    @property
+    def best_value(self):
+        return self.best_distance
+
+    @property
+    def result(self):
+        return self.best_route
+
+    def to_dict(self):
+        return asdict(self)
+
+
+class ACOOptimizer:
+    def __init__(
+        self,
+        *,
+        user_id=None,
+        sdk_code=None,
+        endpoint=None,
+        license_path=None,
+        timeout=30.0,
+        report_usage=True,
+        validation_mode="cached",
+        ants=30,
+        evaporation=0.5,
+        alpha=1.0,
+        beta=2.0,
+        deposit=100.0,
+        iterations=100,
+        seed=None,
+    ):
+        self.endpoint, self.license_user_id, self._usage_reporter = validated_license(
+            endpoint, user_id, sdk_code, license_path, timeout, report_usage, validation_mode
+        )
+        self.ants = int(ants)
+        self.evaporation = float(evaporation)
+        self.alpha = float(alpha)
+        self.beta = float(beta)
+        self.deposit = float(deposit)
+        self.iterations = int(iterations)
+        self.seed = seed
+        if self.ants < 1 or self.iterations < 1:
+            raise ValueError("ants and iterations must be positive integers")
+        if not 0.0 < self.evaporation < 1.0:
+            raise ValueError("evaporation must be between 0 and 1")
+        if self.alpha < 0 or self.beta < 0 or self.deposit <= 0:
+            raise ValueError("alpha and beta cannot be negative; deposit must be positive")
+
+    def solve(self, distance_matrix, *, start_city=0, iterations=None, seed=None):
+        return self.solve_tsp(
+            distance_matrix, start_city=start_city, iterations=iterations, seed=seed
+        )
+
+    @tracked("aco")
+    def solve_tsp(self, distance_matrix, *, start_city=0, iterations=None, seed=None):
+        matrix = square_matrix(distance_matrix, "distance_matrix")
+        size = len(matrix)
+        count = self.iterations if iterations is None else int(iterations)
+        if size < 2:
+            raise ValueError("distance_matrix must contain at least two cities")
+        if any(value < 0 for row in matrix for value in row):
+            raise ValueError("distance_matrix entries cannot be negative")
+        if not isinstance(start_city, int) or not 0 <= start_city < size:
+            raise ValueError("start_city must be a valid city index")
+        if count < 1:
+            raise ValueError("iterations must be a positive integer")
+
+        started = time.perf_counter()
+        rng = random.Random(self.seed if seed is None else seed)
+        pheromone = [[1.0] * size for _ in range(size)]
+        best_route, best_distance = None, math.inf
+        history = []
+
+        def construct_tour():
+            route = [start_city]
+            unvisited = set(range(size))
+            unvisited.remove(start_city)
+            while unvisited:
+                current = route[-1]
+                candidates = sorted(unvisited)
+                weights = [
+                    pheromone[current][city] ** self.alpha
+                    * (1.0 / max(matrix[current][city], 1e-12)) ** self.beta
+                    for city in candidates
+                ]
+                destination = rng.choices(candidates, weights=weights)[0]
+                route.append(destination)
+                unvisited.remove(destination)
+            route.append(start_city)
+            distance = sum(matrix[route[index]][route[index + 1]] for index in range(size))
+            return route, distance
+
+        for _ in range(count):
+            tours = [construct_tour() for _ in range(self.ants)]
+            for route, distance in tours:
+                if distance < best_distance:
+                    best_route, best_distance = list(route), distance
+            for source in range(size):
+                for destination in range(size):
+                    pheromone[source][destination] *= 1.0 - self.evaporation
+            for route, distance in tours:
+                reward = self.deposit / max(distance, 1e-12)
+                for index in range(size):
+                    pheromone[route[index]][route[index + 1]] += reward
+            history.append(float(best_distance))
+
+        return ACOResult(
+            best_route=best_route or [start_city, start_city],
+            best_distance=float(best_distance),
+            history=history,
+            iterations=count,
+            runtime_seconds=time.perf_counter() - started,
+        )
+
+
+def register():
+    return register_algorithm(
+        category="combinatorial",
+        exports={"ACOOptimizer": ACOOptimizer, "ACOResult": ACOResult},
+        specification=AlgorithmSpec(
+            name="aco",
+            family="metaheuristic",
+            problem_types=("tsp",),
+            input_kind="distance_matrix",
+            output_kind="route",
+            factory=ACOOptimizer,
+            description="Ant colony optimization for travelling-salesperson problems.",
+        ),
+    )
+
+
+__all__ = ["ACOOptimizer", "ACOResult", "register"]

ascent_science_aco-0.1.0/src/ascent_science_aco.egg-info/PKG-INFO

@@ -0,0 +1,13 @@
+Metadata-Version: 2.4
+Name: ascent-science-aco
+Version: 0.1.0
+Summary: Ant colony optimizer plugin for Ascent Science SDK
+License-Expression: LicenseRef-Proprietary
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: ascent-science<0.10,>=0.9
+
+# Ascent Science ACO plugin
+
+Install with `python -m pip install ascent-science-aco`, then use
+`ascent.combinatorial.ACOOptimizer(...)`.

ascent_science_aco-0.1.0/src/ascent_science_aco.egg-info/SOURCES.txt

@@ -0,0 +1,9 @@
+README.md
+pyproject.toml
+src/ascent_science_aco/__init__.py
+src/ascent_science_aco.egg-info/PKG-INFO
+src/ascent_science_aco.egg-info/SOURCES.txt
+src/ascent_science_aco.egg-info/dependency_links.txt
+src/ascent_science_aco.egg-info/entry_points.txt
+src/ascent_science_aco.egg-info/requires.txt
+src/ascent_science_aco.egg-info/top_level.txt

ascent_science_aco-0.1.0/src/ascent_science_aco.egg-info/dependency_links.txt

@@ -0,0 +1 @@
+

ascent_science_aco-0.1.0/src/ascent_science_aco.egg-info/entry_points.txt

@@ -0,0 +1,2 @@
+[ascent.algorithms]
+aco = ascent_science_aco:register

ascent_science_aco-0.1.0/src/ascent_science_aco.egg-info/requires.txt

@@ -0,0 +1 @@
+ascent-science<0.10,>=0.9

ascent_science_aco-0.1.0/src/ascent_science_aco.egg-info/top_level.txt

@@ -0,0 +1 @@
+ascent_science_aco