Feat/replace random forest

.github/workflows/pytest.yml

@@ -49,7 +49,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.8", "3.9", "3.10"]
+        python-version: ["3.9", "3.10", "3.11"]
         os: ["ubuntu-latest"]
 
     steps:
@@ -103,7 +103,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.8", "3.9", "3.10"]
+        python-version: ["3.9", "3.10", "3.11"]
         os: ["ubuntu-latest"]
 
     steps:
@@ -143,7 +143,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        python-version: ["3.8", "3.9", "3.10"]
+        python-version: ["3.9", "3.10", "3.11"]
         os: ["ubuntu-latest"]
 
     steps:

CHANGELOG.md

@@ -1,3 +1,7 @@
+# 2.4.0
+## Improvements
+- Replace random forest from pyrfr with random forest from sklearn (#1246)
+
 # 2.3.1
 
 ## Bugfixes

README.md

@@ -47,11 +47,6 @@ conda create -n SMAC python=3.10
 conda activate SMAC
 ```
 
-Install swig:
-```
-conda install gxx_linux-64 gcc_linux-64 swig
-```
-
 Install SMAC via PyPI:
 ```
 pip install smac
@@ -63,6 +58,20 @@ git clone https://github.com/automl/SMAC3.git && cd SMAC3
 make install-dev
 ```
 
+## Running SMAC with pyrfr
+starting from 2.4.0, SMAC uses random forest from [sklearn](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.RandomForestRegressor.html)
+instead of random forest from [pyrfr](https://pypi.org/project/pyrfr/) as the default surrogate model for HPO tasks.
+However, you could still use the old pyrfr surrogate model by calling `smac.facade.old.HyperparameterOptimizationRFRFacade`
+and `smac.facade.old.MultiFidelityRFRFacade`
+
+To work with pyrfr, you need to first install gcc, gxx, and swig:
+```
+conda install gxx_linux-64 gcc_linux-64 swig
+```
+then install smac with the pyrfr option:
+```
+pip install smac[pyrfr]
+```
 
 ## Minimal Example
 

setup.py

@@ -21,6 +21,9 @@ def read_file(filepath: str) -> str:
 
 
 extras_require = {
+    "pyrfr": [
+        "pyrfr>=0.9.0",
+    ],
     "dev": [
         "setuptools",
         "types-setuptools",
@@ -79,8 +82,7 @@ def read_file(filepath: str) -> str:
         "pynisher>=1.0.0",
         "ConfigSpace>=1.0.0",
         "joblib",
-        "scikit-learn>=1.1.2",
-        "pyrfr>=0.9.0",
+        "scikit-learn>=1.6.1",
         "dask[distributed]",
         "dask_jobqueue>=0.8.2",
         "emcee>=3.0.0",

smac/facade/old/__init__.py

@@ -0,0 +1,9 @@
+from smac.facade.old.hyperparameter_optimization_facade_pyrfr import (
+    HyperparameterOptimizationRFRFacade,
+)
+from smac.facade.old.multi_fidelity_facade_pyrfr import MultiFidelityRFRFacade
+
+__all__ = [
+    "HyperparameterOptimizationRFRFacade",
+    "MultiFidelityRFRFacade",
+]

smac/facade/old/hyperparameter_optimization_facade_pyrfr.py

@@ -0,0 +1,50 @@
+from __future__ import annotations
+
+from smac.facade.hyperparameter_optimization_facade import (
+    HyperparameterOptimizationFacade,
+)
+from smac.model.random_forest.pyrfr.random_forest_pyrfr import PyrfrRandomForest
+from smac.scenario import Scenario
+
+
+class HyperparameterOptimizationRFRFacade(HyperparameterOptimizationFacade):
+    @staticmethod
+    def get_model(  # type: ignore
+        scenario: Scenario,
+        *,
+        n_trees: int = 10,
+        ratio_features: float = 1.0,
+        min_samples_split: int = 2,
+        min_samples_leaf: int = 1,
+        max_depth: int = 2**20,
+        bootstrapping: bool = True,
+    ) -> PyrfrRandomForest:
+        """Returns a random forest as surrogate model.
+
+        Parameters
+        ----------
+        n_trees : int, defaults to 10
+            The number of trees in the random forest.
+        ratio_features : float, defaults to 5.0 / 6.0
+            The ratio of features that are considered for splitting.
+        min_samples_split : int, defaults to 3
+            The minimum number of data points to perform a split.
+        min_samples_leaf : int, defaults to 3
+            The minimum number of data points in a leaf.
+        max_depth : int, defaults to 20
+            The maximum depth of a single tree.
+        bootstrapping : bool, defaults to True
+            Enables bootstrapping.
+        """
+        return PyrfrRandomForest(
+            log_y=True,
+            n_trees=n_trees,
+            bootstrapping=bootstrapping,
+            ratio_features=ratio_features,
+            min_samples_split=min_samples_split,
+            min_samples_leaf=min_samples_leaf,
+            max_depth=max_depth,
+            configspace=scenario.configspace,
+            instance_features=scenario.instance_features,
+            seed=scenario.seed,
+        )

smac/facade/old/multi_fidelity_facade_pyrfr.py

@@ -0,0 +1,13 @@
+from __future__ import annotations
+
+from smac.facade.multi_fidelity_facade import MultiFidelityFacade
+from smac.facade.old.hyperparameter_optimization_facade_pyrfr import (
+    HyperparameterOptimizationRFRFacade,
+)
+
+__copyright__ = "Copyright 2022, automl.org"
+__license__ = "3-clause BSD"
+
+
+class MultiFidelityRFRFacade(MultiFidelityFacade, HyperparameterOptimizationRFRFacade):
+    pass