Merge pull request #146 from alan-turing-institute/rename

Rename `grid_search` and `param_grid`

autoemulate/compare.py

@@ -32,9 +32,9 @@ def setup(
         self,
         X,
         y,
-        use_grid_search=False,
-        grid_search_type="random",
-        grid_search_iters=20,
+        param_search=False,
+        param_search_type="random",
+        param_search_iters=20,
         scale=True,
         scaler=StandardScaler(),
         reduce_dim=False,
@@ -53,13 +53,13 @@ def setup(
             Simulation input.
         y : array-like, shape (n_samples, n_outputs)
             Simulation output.
-        use_grid_search : bool
+        param_search : bool
             Whether to perform hyperparameter search over predifined parameter grids.
-        grid_search_type : str
+        param_search_type : str
             Type of hyperparameter search to perform. Can be "grid", "random", or "bayes".
-        grid_search_iters : int
+        param_search_iters : int
             Number of parameter settings that are sampled. Only used if
-            use_grid_search=True and grid_search_type="random".
+            param_search=True and param_search_type="random".
         scale : bool, default=True
             Whether to scale the data before fitting the models using a scaler.
         scaler : sklearn.preprocessing.StandardScaler
@@ -99,9 +99,9 @@ def setup(
         )
         self.metrics = self._get_metrics(METRIC_REGISTRY)
         self.cv = self._get_cv(CV_REGISTRY, fold_strategy, folds)
-        self.use_grid_search = use_grid_search
-        self.search_type = grid_search_type
-        self.grid_search_iters = grid_search_iters
+        self.param_search = param_search
+        self.search_type = param_search_type
+        self.param_search_iters = param_search_iters
         self.scale = scale
         self.scaler = scaler
         self.n_jobs = n_jobs
@@ -185,15 +185,15 @@ def compare(self):
 
         for i in range(len(self.models)):
             # hyperparameter search
-            if self.use_grid_search:
+            if self.param_search:
                 self.models[i] = optimize_params(
                     X=self.X,
                     y=self.y,
                     cv=self.cv,
                     model=self.models[i],
                     search_type=self.search_type,
-                    niter=self.grid_search_iters,
-                    param_grid=None,
+                    niter=self.param_search_iters,
+                    param_space=None,
                     n_jobs=self.n_jobs,
                     logger=self.logger,
                 )

autoemulate/emulators/gaussian_process.py

@@ -68,19 +68,19 @@ def predict(self, X, return_std=False):
 
     def get_grid_params(self, search_type="random"):
         """Returns the grid parameters of the emulator."""
-        param_grid_random = {
+        param_space_random = {
             "nugget": ["fit", "adaptive", "pivot"],
         }
-        param_grid_bayes = {
+        param_space_bayes = {
             "nugget": Categorical(["fit", "adaptive", "pivot"]),
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": False}

autoemulate/emulators/gaussian_process_sk.py

@@ -95,7 +95,7 @@ def predict(self, X, return_std=False):
 
     def get_grid_params(self, search_type="random"):
         """Returns the grid parameters of the emulator."""
-        param_grid_random = {
+        param_space_random = {
             "kernel": [
                 RBF(),
                 Matern(),
@@ -106,19 +106,19 @@ def get_grid_params(self, search_type="random"):
             "alpha": loguniform(1e-10, 1e-2),
             "normalize_y": [True],
         }
-        param_grid_bayes = {
+        param_space_bayes = {
             # "kernel": Categorical([RBF(), Matern()]), # unhashable type
             "optimizer": Categorical(["fmin_l_bfgs_b"]),
             "alpha": Real(1e-10, 1e-2, prior="log-uniform"),
             "normalize_y": Categorical([True]),
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": True}

autoemulate/emulators/gradient_boosting.py

@@ -101,7 +101,7 @@ def predict(self, X):
 
     def get_grid_params(self, search_type="random"):
         """Returns the grid parameters of the emulator."""
-        param_grid_random = {
+        param_space_random = {
             "learning_rate": loguniform(0.01, 0.2),
             "n_estimators": randint(100, 500),
             "max_depth": randint(3, 8),
@@ -112,7 +112,7 @@ def get_grid_params(self, search_type="random"):
             "ccp_alpha": loguniform(0.01, 0.1),
         }
 
-        param_grid_bayes = {
+        param_space_bayes = {
             "learning_rate": Real(0.01, 0.2, prior="log-uniform"),
             "n_estimators": Integer(100, 500),
             "max_depth": Integer(3, 8),
@@ -124,11 +124,11 @@ def get_grid_params(self, search_type="random"):
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": False}

autoemulate/emulators/neural_net_sk.py

@@ -98,7 +98,7 @@ def predict(self, X):
 
     def get_grid_params(self, search_type="random"):
         """Returns the grid parameters of the emulator."""
-        param_grid_random = {
+        param_space_random = {
             "hidden_layer_sizes": [
                 (50,),
                 (100,),
@@ -112,7 +112,7 @@ def get_grid_params(self, search_type="random"):
             "learning_rate_init": loguniform(1e-4, 1e-2),
         }
 
-        param_grid_bayes = {
+        param_space_bayes = {
             # doesn't work with bayes
             # "hidden_layer_sizes": Categorical([
             #     (50,),
@@ -128,11 +128,11 @@ def get_grid_params(self, search_type="random"):
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": True}

autoemulate/emulators/neural_net_torch.py

@@ -141,7 +141,7 @@ def initialize_module(self, reason=None):
         return self
 
     def get_grid_params(self, search_type="random"):
-        param_grid_random = {
+        param_space_random = {
             "lr": loguniform(1e-4, 1e-2),
             "max_epochs": [10, 20, 30],
             "module__hidden_sizes": [
@@ -153,17 +153,17 @@ def get_grid_params(self, search_type="random"):
             ],
         }
 
-        param_grid_bayes = {
+        param_space_bayes = {
             "lr": Real(1e-4, 1e-2, prior="log-uniform"),
             "max_epochs": Integer(10, 30),
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def __sklearn_is_fitted__(self):
         return hasattr(self, "n_features_in_")

autoemulate/emulators/polynomials.py

@@ -82,11 +82,11 @@ def get_grid_params(self, search_type="random"):
             The parameter grid for the model.
         """
         if search_type == "random":
-            param_grid = {}
+            param_space = {}
         elif search_type == "bayes":
-            param_grid = [({"degree": Categorical([2])}, 1)]
+            param_space = [({"degree": Categorical([2])}, 1)]
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": True}

autoemulate/emulators/random_forest.py

@@ -96,7 +96,7 @@ def predict(self, X):
     def get_grid_params(self, search_type="random"):
         """Returns the grid parameters of the emulator."""
 
-        param_grid_random = {
+        param_space_random = {
             "n_estimators": randint(50, 500),
             "min_samples_split": randint(2, 20),
             "min_samples_leaf": randint(1, 10),
@@ -107,7 +107,7 @@ def get_grid_params(self, search_type="random"):
             "max_samples": [None, 0.5, 0.75],
         }
 
-        param_grid_bayes = {
+        param_space_bayes = {
             "n_estimators": Integer(50, 500),
             "min_samples_split": Integer(2, 20),
             "min_samples_leaf": Integer(1, 10),
@@ -119,11 +119,11 @@ def get_grid_params(self, search_type="random"):
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": True}

autoemulate/emulators/rbf.py

@@ -85,13 +85,13 @@ def predict(self, X):
 
     def get_grid_params(self, search_type="random"):
         """Returns the grid parameters of the emulator."""
-        # param_grid_random = {
+        # param_space_random = {
         #     #"smoothing": uniform(0.0, 1.0),
         #     "kernel": ["linear", "thin_plate_spline", "cubic", "quintic", "multiquadric", "inverse_multiquadric", "gaussian"],
         #     #"epsilon": uniform(0.0, 1.0),
         #     "degree": randint(0, 5),
         # }
-        param_grid_random = [
+        param_space_random = [
             {
                 "kernel": ["linear", "multiquadric"],
                 "degree": randint(0, 3),  # Degrees valid for these kernels
@@ -114,7 +114,7 @@ def get_grid_params(self, search_type="random"):
             },
         ]
 
-        param_grid_bayes = [
+        param_space_bayes = [
             {
                 "kernel": Categorical(["linear", "multiquadric"]),
                 "degree": Integer(0, 4),  # Degrees valid for these kernels
@@ -138,11 +138,11 @@ def get_grid_params(self, search_type="random"):
         ]
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": True}

autoemulate/emulators/support_vector_machines.py

@@ -125,7 +125,7 @@ def predict(self, X):
 
     def get_grid_params(self, search_type="random"):
         """Returns the grid paramaters for the emulator."""
-        param_grid_random = {
+        param_space_random = {
             "kernel": ["rbf", "linear", "poly", "sigmoid"],
             "degree": randint(2, 6),
             "gamma": ["scale", "auto"],
@@ -139,7 +139,7 @@ def get_grid_params(self, search_type="random"):
             "max_iter": [-1],
         }
 
-        param_grid_bayes = {
+        param_space_bayes = {
             "kernel": Categorical(["rbf", "linear", "poly", "sigmoid"]),
             "degree": Integer(2, 5),
             "gamma": Categorical(["scale", "auto"]),
@@ -154,11 +154,11 @@ def get_grid_params(self, search_type="random"):
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": False}

autoemulate/emulators/xgboost.py

@@ -132,7 +132,7 @@ def predict(self, X):
 
     def get_grid_params(self, search_type="random"):
         """Returns the grid parameters of the emulator."""
-        param_grid_random = {
+        param_space_random = {
             "booster": ["gbtree", "dart"],
             "n_estimators": randint(100, 1000),
             "max_depth": randint(3, 10),
@@ -148,7 +148,7 @@ def get_grid_params(self, search_type="random"):
             "reg_lambda": loguniform(0.01, 1),
         }
 
-        param_grid_bayes = {
+        param_space_bayes = {
             "booster": Categorical(["gbtree", "dart"]),
             "n_estimators": Integer(100, 1000),
             "max_depth": Integer(3, 10),
@@ -165,11 +165,11 @@ def get_grid_params(self, search_type="random"):
         }
 
         if search_type == "random":
-            param_grid = param_grid_random
+            param_space = param_space_random
         elif search_type == "bayes":
-            param_grid = param_grid_bayes
+            param_space = param_space_bayes
 
-        return param_grid
+        return param_space
 
     def _more_tags(self):
         return {"multioutput": True}