Add be_build_grid and fix some scripts issues

benchmark/Arguments.py

@@ -26,9 +26,11 @@ class EnvData:
 
 class EnvDefault(argparse.Action):
     # Thanks to https://stackoverflow.com/users/445507/russell-heilling
-    def __init__(self, envvar, required=True, default=None, **kwargs):
+    def __init__(
+        self, envvar, required=True, default=None, mandatory=False, **kwargs
+    ):
         self._args = EnvData.load()
-        if required:
+        if required and not mandatory:
             default = self._args[envvar]
             required = False
         super(EnvDefault, self).__init__(

benchmark/Experiments.py

@@ -144,6 +144,7 @@ class BestResults:
             score=self.score_name, model=self.model
         )
         all_files = sorted(list(os.walk(Folders.results)))
+        found = False
         for root, _, files in tqdm(
             all_files, desc="files", disable=self.quiet
         ):
@@ -153,6 +154,9 @@ class BestResults:
                     with open(file_name) as fp:
                         data = json.load(fp)
                     self._process_datafile(results, data, name)
+                    found = True
+        if not found:
+            raise ValueError(f"No results found")
         # Build best results json file
         output = {}
         datasets = Datasets()
@@ -374,10 +378,6 @@ class GridSearch:
         self.grid_file = os.path.join(
             Folders.results, Files.grid_input(score_name, model_name)
         )
-        with open(self.grid_file) as f:
-            self.grid = json.load(f)
-        self.duration = 0
-        self._init_data()
 
     def get_output_file(self):
         return self.output_file
@@ -426,6 +426,10 @@ class GridSearch:
         self.results[name] = [score, hyperparameters, message]
 
     def do_gridsearch(self):
+        with open(self.grid_file) as f:
+            self.grid = json.load(f)
+        self.duration = 0
+        self._init_data()
         now = time.time()
         loop = tqdm(
             list(self.datasets),

benchmark/Results.py

@@ -1124,7 +1124,7 @@ class Summary:
             color2 = TextColor.LINE2
         print(color1, end="")
         print(
-            f"{'Date':10s} {'File':{max_file}s} {'Score':7s} {'Time(h)':7s} "
+            f"{'Date':10s} {'File':{max_file}s} {'Score':8s} {'Time(h)':7s} "
             f"{'Title':s}"
         )
         print(

benchmark/scripts/be_build_best.py

@@ -13,7 +13,11 @@ def main():
     args = arguments.parse()
     datasets = Datasets()
     best = BestResults(args.score, args.model, datasets)
-    best.build()
+    try:
-    if args.report:
+        best.build()
-        report = ReportBest(args.score, args.model, best=True, grid=False)
+    except ValueError:
-        report.report()
+        print("** No results found **")
+    else:
+        if args.report:
+            report = ReportBest(args.score, args.model, best=True, grid=False)
+            report.report()

benchmark/scripts/be_build_grid.py

@@ -2,9 +2,17 @@
 import os
 import json
 from benchmark.Utils import Files, Folders
+from benchmark.Arguments import Arguments
+
+"""Build sample grid input file for the model with data taken from the
+input grid used optimizing STree
+"""
 
 
 def main():
+    arguments = Arguments()
+    arguments.xset("model", mandatory=True).xset("score", mandatory=True)
+    args = arguments.parse()
     data = [
         '{"C": 1e4, "gamma": 0.1, "kernel": "rbf"}',
         '{"C": 7, "gamma": 0.14, "kernel": "rbf"}',
@@ -105,8 +113,8 @@ def main():
         output.append(results_tmp)
 
     # save results
-    file_name = Files.grid_input("accuracy", "ODTE")
+    file_name = Files.grid_input(args.score, args.model)
     file_output = os.path.join(Folders.results, file_name)
     with open(file_output, "w") as f:
         json.dump(output, f, indent=4)
-    print(f"Grid values saved to {file_output}")
+    print(f"Generated grid input file to {file_output}")

benchmark/scripts/be_grid.py

@@ -8,9 +8,11 @@ from benchmark.Arguments import Arguments
 
 def main():
     arguments = Arguments()
-    arguments.xset("score").xset("platform").xset("model").xset("n_folds")
+    arguments.xset("score").xset("platform").xset("model")
-    arguments.xset("quiet").xset("stratified").xset("dataset")
+    arguments.xset("quiet").xset("stratified").xset("dataset").xset("n_folds")
     args = arguments.parse()
+    if not args.quiet:
+        print(f"Perform grid search with {args.model} model")
     job = GridSearch(
         score_name=args.score,
         model_name=args.model,
@@ -18,6 +20,10 @@ def main():
         datasets=Datasets(dataset_name=args.dataset),
         progress_bar=not args.quiet,
         platform=args.platform,
-        folds=args.folds,
+        folds=args.n_folds,
     )
-    job.do_gridsearch()
+    try:
+        job.do_gridsearch()
+    except FileNotFoundError:
+        print(f"** The grid input file [{job.grid_file}] could not be found")
+        print("")

setup.py

@@ -67,6 +67,7 @@ setuptools.setup(
             "be_benchmark=benchmark.scripts.be_benchmark:main",
             "be_best=benchmark.scripts.be_best:main",
             "be_build_best=benchmark.scripts.be_build_best:main",
+            "be_build_grid=benchmark.scripts.be_build_grid:main",
             "be_grid=benchmark.scripts.be_grid:main",
             "be_pair_check=benchmark.scripts.be_pair_check:main",
             "be_print_strees=benchmark.scripts.be_print_strees:main",