Merge pull request #36 from gugarosa/feat/walrus_optimization_algorithm

gugarosa · web-flow · commit 19264745d0ef · 2023-09-22T07:41:50.000-07:00
chore(opytimizer): Adds Walrus Optimization Algorithm
diff --git a/examples/optimizers/swarm/create_waoa.py b/examples/optimizers/swarm/create_waoa.py
@@ -0,0 +1,4 @@
+from opytimizer.optimizers.swarm import WAOA
+
+# Creates a WAOA optimizer
+o = WAOA()
diff --git a/opytimizer/optimizers/swarm/__init__.py b/opytimizer/optimizers/swarm/__init__.py
@@ -29,4 +29,5 @@
 from opytimizer.optimizers.swarm.ssa import SSA
 from opytimizer.optimizers.swarm.sso import SSO
 from opytimizer.optimizers.swarm.stoa import STOA
+from opytimizer.optimizers.swarm.waoa import WAOA
 from opytimizer.optimizers.swarm.woa import WOA
diff --git a/opytimizer/optimizers/swarm/waoa.py b/opytimizer/optimizers/swarm/waoa.py
@@ -0,0 +1,108 @@
+"""Walrus Optimization Algorithm.
+"""
+
+import copy
+from typing import Any, Dict, Optional
+
+import numpy as np
+
+import opytimizer.math.random as r
+import opytimizer.utils.logging as l
+from opytimizer.core.optimizer import Optimizer
+from opytimizer.core.function import Function
+from opytimizer.core.space import Space
+
+logger = l.get_logger(__name__)
+
+class WAOA(Optimizer):
+    """A WAOA class, inherited from Optimizer.
+
+    This is the designed class to dife WAOA-related 
+    variables and methods.
+
+    References:
+        P. Trojovsk&yacute; and M. Dehghani. A new bio-inspired metaheuristic algorithm for
+        solving optimization problems based on walruses behavior. Scientific Reports (2023).
+
+    """
+
+    def __init__(self, params: Optional[Dict[str, Any]] = None) -> None:
+        """Initialization method.
+
+        Args:
+            params (str): Contains key-value parameters to the meta-heuristics.
+        """
+
+        logger.info('Overriding class: Optimizer -> SSA')
+        
+        super(WAOA, self).__init__()
+
+        self.build(params)
+
+        logger.info('Class overrided.')
+
+    def update(self, space: Space, function: Function, iteration: int) -> None:
+        """Wraps Walrus Optimization Algorithm over all agents and variables.
+
+        Args:
+            space: Space containing agents and update-related information.
+            function: A Function object that will be used as the objective function.
+            iteration: Current iteration.
+
+        """
+
+        for i, agent in enumerate(space.agents):
+            a = copy.deepcopy(agent)
+
+            for j in range(space.n_variables):
+                
+                r1 = r.generate_integer_random_number(1, 2)
+                r2 = r.generate_uniform_random_number()
+
+                a.position[j] = agent.position[j] + r2 * (space.best_agent.position[j] - r1 * agent.position[j])
+
+            a.clip_by_bound()
+
+            a.fit = function(a.position)
+            if a.fit < agent.fit:
+                agent.position = copy.deepcopy(a.position)
+                agent.fit = copy.deepcopy(a.fit)
+
+            k = r.generate_integer_random_number(0, space.n_agents, i)
+            
+            if (space.agents[k].fit < agent.fit):
+
+                for j in range(space.n_variables):
+                
+                    r1 = r.generate_integer_random_number(1, 2)
+                    r2 = r.generate_uniform_random_number()
+
+                    a.position[j] = agent.position[j] + r2 * (space.agents[k].position[j] - r1 * agent.position[j])
+            
+            else:
+
+                for j in range(space.n_variables):
+                
+                    r2 = r.generate_uniform_random_number()
+
+                    a.position[j] = agent.position[j] + r2 * (agent.position[j] - space.agents[k].position[j])
+            
+            a.clip_by_bound()
+
+            a.fit = function(a.position)
+            if a.fit < agent.fit:
+                agent.position = copy.deepcopy(a.position)
+                agent.fit = copy.deepcopy(a.fit)
+
+            for j in range(space.n_variables):
+
+                r2 = r.generate_uniform_random_number()
+
+                a.position[j] = agent.position[j] + ((agent.ub[j]/(iteration+1)) - r2 * (agent.lb[j]/(iteration+1)))
+
+            a.clip_by_bound()
+
+            a.fit = function(a.position)
+            if a.fit < agent.fit:
+                agent.position = copy.deepcopy(a.position)
+                agent.fit = copy.deepcopy(a.fit)
diff --git a/tests/opytimizer/optimizers/swarm/test_waoa.py b/tests/opytimizer/optimizers/swarm/test_waoa.py
@@ -0,0 +1,16 @@
+import numpy as np
+
+from opytimizer.optimizers.swarm import waoa
+from opytimizer.spaces import search
+
+def test_waoa_update():
+    def square(x):
+        return np.sum(x**2)
+
+    new_waoa = waoa.WAOA()
+
+    search_space = search.SearchSpace(
+        n_agents=20, n_variables=2, lower_bound=[0, 0], upper_bound=[10, 10]
+    )
+
+    new_waoa.update(search_space, square)
