Merge pull request #41 from tsukuba-websci/nanami/fix/io

fix & update: 出力形式の変更

ga/.gitignore

@@ -11,3 +11,5 @@ results/twitter/**/*
 
 results/synthetic/**/*
 !results/synthetic/.gitkeep
+
+results/grid_search/**/*

ga/ga.py

@@ -128,10 +128,12 @@ def dump_population(self, population: list, generation: int, fitness: list) -> N
             fitness (list): 適応度
         """
         fp = f"{self.archives_dir}/{str(generation).zfill(8)}.csv"
+        population_fitness = sorted(list(zip(population, fitness)), key=lambda x: -x[1])
+
         with open(fp, "w") as f:
             writer = csv.writer(f)
-            writer.writerow(["rho", "nu", "recentness", "frequency", "objective"])
-            for individual, fit in zip(population, fitness):
+            writer.writerow(["rho", "nu", "recentness", "frequency", "distance"])
+            for individual, fit in population_fitness:
                 writer.writerow([individual[0], individual[1], individual[2], individual[3], -1 * fit])
 
     def plot():

ga/grid_search.py

@@ -131,7 +131,7 @@ def main():
     # Set Up GridSearch
     mutation_rate_iter = [0.01, 0.02, 0.03, 0.04, 0.05]
     cross_rate_iter = [0.8, 0.85, 0.9, 0.95]
-    population_size_iter = [20, 40, 60, 80, 100]
+    population_size_iter = [10, 20, 30, 40, 50]
     num_generations = 100
     output_dir = f"./results/grid_search/{target_data}"
     os.makedirs(output_dir, exist_ok=True)

ga/io_utils.py

@@ -33,17 +33,19 @@ def parse_args(parser: argparse.ArgumentParser) -> argparse.Namespace:
     return args
 
 
-def export_individual(distance: float, individual: list, fpath: str) -> None:
+def export_individual(
+    distance: float, individual: list, population_size: int, mutation_rate: float, cross_rate: float, fpath: str
+) -> None:
     """個体をCSVファイルに出力する．
 
     Args:
         distance (float): ターゲットとの距離
         individual (list): 個体を表すタプル．(rho, nu, recentness, frequency)の順
         fpath (str): 出力先のパス
     """
-    header = ["rho", "nu", "recentness", "frequency", "objective"]
+    header = ["rho", "nu", "recentness", "frequency", "objective", "populasion_size", "mutation_rate", "cross_rate"]
     objective = distance
-    row = [*individual, objective]
+    row = [*individual, objective, population_size, mutation_rate, cross_rate]
     with open(fpath, "w") as f:
         writer = csv.writer(f)
         writer.writerow(header)

ga/main.py

@@ -152,7 +152,7 @@ def main():
         archive_dir=os.path.join(output_base_dir, "archives"),
     )
 
-    export_individual(min_distance, best_individual, output_fp)
+    export_individual(min_distance, best_individual, population_size, mutation_rate, cross_rate, output_fp)
 
     logging.info(f"Finihsed GA. Result is dumped to {target_data}")
 

qd/main.py

@@ -47,6 +47,7 @@ def __init__(
 
     def run(self):
         self.jl_main, self.thread_num = JuliaInitializer().initialize()
+        history2vec_ = History2Vec(self.jl_main, self.thread_num)
 
         archive: Union[CVTArchive, None] = None
         if os.path.exists(f"{self.archives_dir_path}/archive.pkl"):
@@ -105,7 +106,7 @@ def run(self):
             with Pool(self.thread_num) as pool:
                 histories = pool.map(run_model, params_list)
 
-            history_vecs = History2Vec(self.jl_main, self.thread_num).history2vec_parallel(histories, 1000)
+            history_vecs = history2vec_.history2vec_parallel(histories, 1000)
 
             bcs = self.history2bd.run(histories)
 
@@ -131,7 +132,9 @@ def run(self):
                 },
                 inplace=True,
             )
-            df = df[["rho", "nu", "recentness", "frequency", "objective"]].sort_values(by="objective", ascending=False)
+            df["objective"] = -df["objective"]
+            df.rename(columns={"objective": "distance"}, inplace=True)
+            df = df[["rho", "nu", "recentness", "frequency", "distance"]].sort_values(by="distance", ascending=True)
             df.to_csv(f"{self.archives_dir_path}/{iter:0>8}.csv", index=False)
 
             if iter % 25 == 0:
@@ -141,7 +144,8 @@ def run(self):
                 assert archive.stats is not None, "archive.stats is None!"
                 print(f"  - Max Score: {archive.stats.obj_max}")
         # save best result as csv
-        df.head(1).to_csv(f"{self.result_dir_path}/best.csv", index=False)
+        if not os.path.exists(f"{self.result_dir_path}/best.csv"):
+            df.head(1).to_csv(f"{self.result_dir_path}/best.csv", index=False)
 
 
 if __name__ == "__main__":