revised the evaluators documentation

deepcave/evaluators/__init__.py

@@ -5,12 +5,9 @@
 ## Modules
 ----------
 fanova
-    This module provides a tool for assessing the importance of an algorithms hyperparameters (HPs).
+    This module provides a tool for assessing the importance of an algorithms Hyperparameters.
 footprint
-    This module provides utilities for creating a footprint of a model.
+    This module provides utilities for creating a footprint of a run.
 lpi
     This module provides utilties to calculate the local parameter importance (LPM).
-
-## Packages
-epm
 """

deepcave/evaluators/epm/__init__.py

@@ -5,11 +5,12 @@
 ## Modules
 ----------
 fanova_forest
-    This module provides utilities for conducting feature analysis.
+    The module provides utilities for creating a fANOVA forest.
 random_forest
-    This module is used for training and using a Random Forest Regression model.
+    This module can be used for training and using a Random Forest Regression model.
 random_forest_surrogate
-    This module provides an implementation for a RandomForest Surrogate model.
+    This module provides a RandomForest Surrogate model.
 utils
-    Return types and bounds of hyperparameter (HP).
+    This module provides a utility to get the types 
+    as well as the bounds of the Hyperparameters.
 """

deepcave/evaluators/epm/fanova_forest.py

@@ -2,14 +2,13 @@
 """
 # FanovaForest
 
-The module provides utilities for conducting feature analysis.
+The module provides utilities for creating a fANOVA forest.
 
-It includes a FanovaForest wrapper for pyrfr, which facilitates this analysis.
-The functionality of a FanovaForest Regression model is wrapped.
-FanovaForest can be used for analyzing and quantifying the features of a dataset.
+It includes a FanovaForest wrapper for pyrfr.
+fANOVA can be used for analyzing the importances of Hyperparameters.
 
 ## Classes
-    - FanovaForest: A fanova forest wrapper for pyrfr.
+    - FanovaForest: A fANOVA forest wrapper for pyrfr.
 """
 
 from typing import Any, Dict, List, Optional, Tuple, Union
@@ -27,21 +26,20 @@
 
 class FanovaForest(RandomForest):
     """
-    A fanova forest wrapper for pyrfr.
+    A fANOVA forest wrapper for pyrfr.
 
     Properties
     ----------
     cutoffs : Tuple[float, float]
-        The cutoff of the model.
+        The cutoffs of the model.
     percentiles : NDArray[floating]
         The percentiles of the data points Y.
     all_midpoints : List
         All midpoints tree wise for the whole forest.
     all_sizes : List
-        All interval sized tree wise for the whole forest.
+        All interval sizes tree wise for the whole forest.
     bounds : List[Tuple[float, float]
         Stores feature bounds.
-        Can be used as categorical or continuous parameter.
     trees_total_variances : List
         The total variances of the trees.
     trees_total_variance : Any
@@ -53,7 +51,7 @@ class FanovaForest(RandomForest):
     V_U_individual : Dict[Tuple[int, ...], List[Any]]
         Store variance-related information for individual subsets.
     n_params : int
-        The number of parameters to sample.
+        The number of Hyperparameters to sample.
     """
 
     def __init__(
@@ -167,12 +165,13 @@ def compute_marginals(
         self, hp_ids: Union[List[int], Tuple[int, ...]], depth: int = 1
     ) -> Tuple[Dict[Tuple[int, ...], List[Any]], Dict[Tuple[int, ...], List[Any]],]:
         """
-        Return the marginal of selected parameters.
+        Return the marginal of selected Hyperparameters.
 
         Parameters
         ----------
         hp_ids: Union[List[int], Tuple[int, ...]]
-            Contains the indices of the configspace for the selected parameters (starts with 0).
+            Contains the indices of the configspace for the selected Hyperparameters 
+            (starts with 0).
         depth: int
             The depth of the marginalization.
             Default value is 1.
@@ -181,7 +180,7 @@ def compute_marginals(
         -------
         Tuple[Dict[Tuple[int, ...], List[Any]],
         Dict[Tuple[int, ...], List[Any]],
-            The marginal of selected parameters.
+            The marginal of selected Hyperparameters.
         """
         if not isinstance(hp_ids, tuple):
             hp_ids = tuple(hp_ids)

deepcave/evaluators/epm/random_forest.py

@@ -2,7 +2,7 @@
 """
 # RandomForest
 
-This module is used for training and using a Random Forest Regression model.
+This module can be used for training and using a Random Forest Regression model.
 
 A pyrfr wrapper is used for simplification.
 
@@ -12,7 +12,6 @@
 ## Constants
     VERY_SMALL_NUMBER : float
     PYRFR_MAPPING : Dict[str, str]
-    }
 """
 
 from typing import Any, Dict, List, Optional, Tuple, Union
@@ -68,11 +67,11 @@ class RandomForest:
     seed : int
         The seed. If not provided, it is random.
     types : List[int]
-        The types of the hyperparameters (HPS).
+        The types of the Hyperparameters.
     bounds : List[Tuple[float, float]]
-        The bounds of the hyperparameters (HPS).
+        The bounds of the Hyperparameters.
     n_params : int
-        The number of hyperparameters (HPS) in the configuration space.
+        The number of Hyperparameters in the configuration space.
     n_features : int
         The number of features.
     pca_components : int
@@ -202,7 +201,7 @@ def _impute_inactive(self, X: np.ndarray) -> np.ndarray:
         Raises
         ------
         ValueError
-            If hyperparameter (HP) is not supported.
+            If Hyperparameter is not supported.
         """
         conditional: Dict[int, bool] = {}
         impute_values: Dict[int, float] = {}
@@ -296,7 +295,7 @@ def train(self, X: np.ndarray, Y: np.ndarray) -> None:
         """
         Train the random forest on X and Y.
 
-        Transforms X if principal component analysis (PCA) is applied.
+        Transform X if principal component analysis (PCA) is applied.
         Afterwards, `_train` is called.
 
         Parameters
@@ -403,7 +402,8 @@ def _predict(self, X: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
 
         Parameters
         ----------
-        X : np.ndarray [n_samples, n_features (config + instance features)]
+        X : np.ndarray 
+            [n_samples, n_features (config + instance features)]
 
         Returns
         -------
@@ -457,7 +457,7 @@ def predict_marginalized(self, X: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
         """
         Predict mean and variance marginalized over all instances.
 
-        Returns the predictive mean and variance marginalized over all
+        Return the predictive mean and variance marginalized over all
         instances for a set of configurations.
 
         Parameters
@@ -525,7 +525,7 @@ def get_leaf_values(self, x: np.ndarray) -> regression.binary_rss_forest:
         Parameters
         ----------
         x : np.ndarray
-            Input data points.
+            Input data array.
 
         Returns
         -------

deepcave/evaluators/epm/random_forest_surrogate.py

@@ -2,7 +2,7 @@
 """
 # RandomForest Surrogate
 
-This module provides an implementation for a RandomForest Surrogate model.
+This module provides a RandomForest Surrogate model.
 
 Mean and standard deviation values can be predicted for a given input with this module.
 

deepcave/evaluators/epm/utils.py

@@ -2,7 +2,8 @@
 """
 # Utils
 
-This module provides a utility to get the types of the hyperparameters (HPs).
+This module provides a utility to get the types 
+as well as the bounds of the Hyperparameters.
 """
 
 import typing
@@ -27,9 +28,9 @@ def get_types(
     instance_features: typing.Optional[np.ndarray] = None,
 ) -> typing.Tuple[typing.List[int], typing.List[typing.Tuple[float, float]]]:
     """
-    Return the types of the hyperparameters (HPs).
+    Return the types of the Hyperparameters.
 
-    Also return the bounds of the hyperparameters (HPs) and instance features.
+    Also return the bounds of the Hyperparameters and instance features.
 
     Parameters
     ----------
@@ -42,14 +43,14 @@ def get_types(
     Returns
     -------
     Tuple[typing.List[int], List[Tuple[float, float]]]
-        The types of the hyperparameters (HPs), as well as the bounds and instance features.
+        The types of the Hyperparameters, as well as the bounds and instance features.
 
     Raises
     ------
     ValueError
-        Inactive parameters not supported for Beta and Normal Hyperparameters (HPs).
+        Inactive parameters not supported for Beta and Normal Hyperparameters.
     TypeError
-        If the hyperparameter (HP) Type is unknown.
+        If the Hyperparameter Type is unknown.
     """
     # Extract types vector for rf from config space and the bounds
     types = [0] * len(config_space.get_hyperparameters())

deepcave/evaluators/fanova.py

@@ -2,7 +2,7 @@
 """
 # fANOVA
 
-This module provides a tool for assessing the importance of an algorithms hyperparameters (HPs).
+This module provides a tool for assessing the importance of an algorithms Hyperparameters.
 
 Utilities provide calculation of the data wrt the budget and train the forest on the encoded data.
 
@@ -35,9 +35,9 @@ class fANOVA:
     cs : ConfigurationSpace
         The configuration space of the run.
     hps : List[Hyperparameters]
-        The hyperparameters (HPs) of the configuration space.
+        The Hyperparameters of the configuration space.
     hp_names : List[str]
-        The corresponding names of the hyperparameters (HPs).
+        The corresponding names of the Hyperparameters.
     n_trees : int
         The number of trees.
     """
@@ -100,7 +100,7 @@ def get_importances(
         self, hp_names: Optional[List[str]] = None, depth: int = 1, sort: bool = True
     ) -> Dict[Union[str, Tuple[str, ...]], Tuple[float, float, float, float]]:
         """
-        Return the importance scores from the passed hyperparameter (HP) names.
+        Return the importance scores from the passed Hyperparameter names.
 
         Warning
         -------
@@ -109,17 +109,17 @@ def get_importances(
         Parameters
         ----------
         hp_names : Optional[List[str]]
-            Selected hyperparameter (HP) names to get the importance scores from. If None, all
-            hyperparameters (HPs) of the configuration space are used.
+            Selected Hyperparameter names to get the importance scores from. If None, all
+            Hyperparameters of the configuration space are used.
         depth : int, optional
             How often dimensions should be combined. By default 1.
         sort : bool, optional
-            Whether the hyperparameters (HPs) should be sorted by importance. By default True.
+            Whether the Hyperparameters should be sorted by importance. By default True.
 
         Returns
         -------
         Dict[Union[str, Tuple[str, ...]], Tuple[float, float, float, float]]
-            Dictionary with hyperparameter (HP) names and the corresponding importance scores.
+            Dictionary with Hyperparameter names and the corresponding importance scores.
             The values are tuples of the form (mean individual, var individual, mean total,
             var total). Note that individual and total are the same if depth is 1.
 

deepcave/evaluators/footprint.py

@@ -2,13 +2,13 @@
 """
 # Footprint
 
-This module provides utilities for creating a footprint of a model.
+This module provides utilities for creating a footprint of a run.
 It uses multidimensional scaling (MDS).
 It also provides utilities to get the surface and the points of the plot.
 
 
 ## Classes
-    - Footprint: Can train and create a footprint of a model.
+    - Footprint: Can train and create a footprint of a run.
 """
 
 from typing import List, Optional, Tuple, Union
@@ -34,7 +34,7 @@
 
 class Footprint:
     """
-    Can train and create a footprint of a model.
+    Can train and create a footprint of a run.
 
     It uses multidimensional scaling (MDS).
     Provides utilities to get the surface and the points of the plot.
@@ -336,7 +336,7 @@ def _get_max_distance(self) -> float:
         """
         Calculate the maximum distance between all configs.
 
-        Basically, the number of hyperparameters (HPS) are just counted.
+        Basically, the number of Hyperparameters are just counted.
 
         Returns
         -------
@@ -514,12 +514,12 @@ def _get_depth(self, hp: Hyperparameter) -> int:
         Parameters
         ----------
         hp: Hyperparameter
-            name of parameter to inspect
+            name of Hyperparameter to inspect
 
         Returns
         -------
         int
-            Depth of the hyperparameter (HP).
+            Depth of the Hyperparameter.
         """
         parents = self.cs.get_parents_of(hp)
         if not parents:
@@ -562,7 +562,7 @@ def _get_mds(self) -> np.ndarray:
 
     def _train_on_objective(self, X: np.ndarray, Y: np.ndarray) -> None:
         """
-        Trains the random forest on the performance.
+        Train the random forest on the performance.
 
         Parameters
         ----------
@@ -577,7 +577,7 @@ def _train_on_objective(self, X: np.ndarray, Y: np.ndarray) -> None:
 
     def _train_on_areas(self) -> None:
         """
-        Trains the random forest on the "valid" areas.
+        Train the random forest on the "valid" areas.
 
         Raises
         ------

deepcave/evaluators/lpi.py

@@ -31,21 +31,21 @@ class LPI:
     Properties
     ----------
     run : AbstractRun
-        The AbstractRun ro calculate the importance from.
+        The AbstractRun to get the importance from.
     cs : ConfigurationSpace
         The configuration space of the run.
     hp_names : List[str]
-        The names of the hyperparameters (HPs).
+        The names of the Hyperparameters.
     variances : Dict[Any, list]
         The overall variances per tree.
     importances : dict
-        The importances of the hyperparameters (HPs).
+        The importances of the Hyperparameters.
     continuous_neighbors : int
-        The number of neighbors chosen for continuous hyperparameters (HPs).
+        The number of neighbors chosen for continuous Hyperparameters.
     incumbent : Configuration
         The incumbent of the run.
     default : Configuration
-        A configuration containing hyperparameters (HPs) with default values.
+        A configuration containing Hyperparameters with default values.
     incumbent_array : numpy.ndarray
         The internal vector representation of the incumbent.
     seed : int
@@ -231,12 +231,12 @@ def get_importances(self, hp_names: List[str]) -> Dict[str, Tuple[float, float]]
         Parameters
         ----------
         hp_names : List[str]
-            Selected hyperparameter (HP) names to get the importance scores from.
+            Selected Hyperparameter names to get the importance scores from.
 
         Returns
         -------
         importances : Dict[str, Tuple[float, float]]
-            Hyperparameter (HP) name and mean+var importance.
+            Hyperparameter name and mean+var importance.
 
         Raises
         ------