Implemented a fix for the non-constrained version of GA

Author: fjwillemsen

kernel_tuner/strategies/genetic_algorithm.py

@@ -36,7 +36,7 @@ def tune(searchspace: Searchspace, runner, tuning_options):
     population = GA.generate_population()
 
     for generation in range(generations):
-        if any([not searchspace.is_param_config_valid(tuple(dna)) for dna in population]):
+        if constraint_aware and any([not searchspace.is_param_config_valid(tuple(dna)) for dna in population]):
             raise ValueError(f"Generation {generation}/{generations}, population validity: {[searchspace.is_param_config_valid(tuple(dna)) for dna in population]}")
 
         # determine fitness of population members
@@ -77,7 +77,7 @@ def tune(searchspace: Searchspace, runner, tuning_options):
             for child in children:
                 child = GA.mutate(child)
 
-                if child not in population and searchspace.is_param_config_valid(tuple(child)):
+                if child not in population and (not constraint_aware or searchspace.is_param_config_valid(tuple(child))):
                     population.append(child)
 
                 if len(population) >= pop_size:
@@ -161,7 +161,7 @@ def mutate(self, dna, cache=False):
                 if cache:
                     neighbors = self.searchspace.get_neighbors(tuple(dna), neighbor_method="Hamming")
                 else:
-                    neighbors = self.searchspace.get_neighbors(tuple(dna), neighbor_method="Hamming")
+                    neighbors = self.searchspace.get_neighbors_no_cache(tuple(dna), neighbor_method="Hamming")
                 if len(neighbors) > 0:
                     return list(random.choice(neighbors))
             else: