Merge branch 'development' into v2.3.1

automl · Jan 21, 2025 · d2cbb7b · d2cbb7b
2 parents d0d988c + 3c36ce0
commit d2cbb7b
Show file tree

Hide file tree

Showing 134 changed files with 546 additions and 289 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -5,12 +5,18 @@
 - Correct intensifier for Algorithm Configuration Facade (#1162, #1165)
 - Migrate sphinx docs to mkdocs (#1155)
 
-# 2.2.1
+## Bugfixes
+- Fix bug in differential evolution acquisition maximizer in case the search space contains categorical hyperparameters (#1150)
+- Fix wrong attr query in the target function runner (crash cost) (#1176)
+- Fix kwargs for DifferentialEvolution (#1187)
+- Fix PiBo implementation (#1076)
+- Add fix for local search running forever (#1194)
 
 ## Improvements
 - Add logger information on handling of stopIteration error (#960)
 - Replace deprecated ConfigSpace methods (#1139)
 - Separated Wallclock time measurements from CPU time measurements and storing them under new 'cpu_time' variable (#1173)
+- Adapt RunHistory to be human readable (# 1174)
 
 ## Dependencies
 - Allow numpy >= 2.x (#1146)

diff --git a/CITATION.cff b/CITATION.cff
@@ -9,7 +9,7 @@ date-released: "2016-08-17"
 url: "https://automl.github.io/SMAC3/master/index.html"
 repository-code: "https://github.com/automl/SMAC3"
 
-version: "2.2.0"
+version: "2.3.0"
 
 type: "software"
 keywords:

diff --git a/LICENSE.txt b/LICENSE.txt
@@ -4,7 +4,7 @@ SMAC License
 
 BSD 3-Clause License
 
-Copyright (c) 2016-2018, Ml4AAD Group (http://www.ml4aad.org/)
+Copyright (c) 2025, Leibniz University Hannover - Institute of AI (https://www.ai.uni-hannover.de/)
 All rights reserved.
 
 Redistribution and use in source and binary forms, with or without

diff --git a/Makefile b/Makefile
@@ -5,7 +5,7 @@ SHELL := /bin/bash
 
 NAME := SMAC3
 PACKAGE_NAME := smac
-VERSION := 2.2.0
+VERSION := 2.3.0
 
 DIR := "${CURDIR}"
 SOURCE_DIR := ${PACKAGE_NAME}

diff --git a/README.md b/README.md
@@ -12,7 +12,7 @@ SMAC offers a robust and flexible framework for Bayesian Optimization to support
 hyperparameter configurations for their (Machine Learning) algorithms, datasets and applications at hand. The main core 
 consists of Bayesian Optimization in combination with an aggressive racing mechanism to efficiently decide which of two configurations performs better.
 
-SMAC3 is written in Python3 and continuously tested with Python 3.8, 3.9, and 3.10. Its Random
+SMAC3 is written in Python3 and continuously tested with Python 3.8, 3.9, and 3.10 (and works with newer python versions). Its Random
 Forest is written in C++. In further texts, SMAC is representatively mentioned for SMAC3.
 
 > [Documentation](automl.github.io/SMAC3/latest/)
@@ -144,4 +144,4 @@ If you use SMAC in one of your research projects, please cite our
 }
 ```
 
-Copyright (C) 2016-2022  [AutoML Group](http://www.automl.org).
+Copyright (c) 2025, [Leibniz University Hannover - Institute of AI](https://www.ai.uni-hannover.de/)
diff --git a/docs/1_installation.md b/docs/1_installation.md
@@ -1,5 +1,4 @@
 # Installation
-
 ## Requirements
 
 SMAC is written in python3 and therefore requires an environment with python>=3.8.

diff --git a/docs/advanced_usage/10_continue.md b/docs/advanced_usage/10_continue.md
@@ -19,4 +19,4 @@ The behavior can be controlled by setting the parameter ``overwrite`` in the fac
     and the old run is not affected.
 
 
-Please have a look at our [continue example](../examples/1%20Basics/5_continue.html).
+Please have a look at our [continue example](../examples/1%20Basics/5_continue.md).
diff --git a/docs/advanced_usage/2_multi_fidelity.md b/docs/advanced_usage/2_multi_fidelity.md
@@ -13,5 +13,11 @@ target function but ``min_budget`` and ``max_budget`` are used internally to det
 each stage. That's also the reason why ``min_budget`` and ``max_budget`` are *not required* when using instances: 
 The ``max_budget`` is simply the max number of instances, whereas the ``min_budget`` is simply 1.
 
+!!! warning
+    ``smac.main.config_selector.ConfigSelector`` contains the ``min_trials`` parameter. This parameter determines
+    how many samples are required to train the surrogate model. If budgets are involved, the highest budgets 
+    are checked first. For example, if min_trials is three, but we find only two trials in the runhistory for
+    the highest budget, we will use trials of a lower budget instead.
+
 Please have a look into our [multi-fidelity examples](Multi-Fidelity and Multi-Instances) to see how to use
 multi-fidelity optimization in real-world applications.
diff --git a/docs/advanced_usage/3_multi_objective.md b/docs/advanced_usage/3_multi_objective.md
@@ -37,4 +37,4 @@ use the method ``get_incumbents`` in the intensifier.
   incumbents = smac.intensifier.get_incumbents()
 ```
 
-We show an example of how to use multi-objective with plots in our [examples](../examples/3%20Multi-Objective/1_schaffer.html).
+We show an example of how to use multi-objective with plots in our [examples](../examples/3%20Multi-Objective/1_schaffer.md).
diff --git a/docs/advanced_usage/5.1_warmstarting.md b/docs/advanced_usage/5.1_warmstarting.md
@@ -32,7 +32,7 @@ class TrialInfo:
 ```
 
 ## Usage Example
-See [`examples/1_basics/8_warmstart.py`](../../examples/1%20Basics/8_warmstart.py).
+See [`examples/1_basics/8_warmstart.py`](../examples/1%20Basics/8_warmstart.md).
 
 
 ```python

diff --git a/docs/advanced_usage/5_ask_and_tell.md b/docs/advanced_usage/5_ask_and_tell.md
@@ -16,4 +16,4 @@ and report the results of the trial.
     into the intensification process.
 
 
-Please have a look at our [ask-and-tell example](../examples/1%20Basics/3_ask_and_tell.html).
+Please have a look at our [ask-and-tell example](../examples/1%20Basics/3_ask_and_tell.md).
diff --git a/docs/advanced_usage/8_logging.md b/docs/advanced_usage/8_logging.md
@@ -38,7 +38,7 @@ Scenario(
     ...
 )
 ```
-Notably, if an output already exists at `./some_directory/experiment_name/seed`, the behavior is determined by the overwrite parameter in the [facade's][smac/facade/abstract_facade] settings. This parameter specifies whether to continue the previous run (default) or start a new run.
+Notably, if an output already exists at `./some_directory/experiment_name/seed`, the behavior is determined by the overwrite parameter in the [facade's][smac.facade.abstract_facade] settings. This parameter specifies whether to continue the previous run (default) or start a new run.
 
 The output is split into four different log files, and a copy of the utilized [Configuration Space of the ConfigSpace library](https://automl.github.io/ConfigSpace/latest/).
 
@@ -99,21 +99,22 @@ The runhistory.json in split into four parts. `stats`, `data`, `configs`, and `c
 `data` contains a list of entries, one for each configuration.
 ```json
   "data": [
-    [
-      1,                            # config_id
-      null,                         # instance or None
-      209652396,                    # seed or None
-      null,                         # budget or None
-      5.4345623938566385,           # cost
-      6.699562072753906e-05,        # time
-      6.299999999992423e-05,        # cpu_time
-      1,                            # status
-      1733133181.2144582,           # start_time
-      1733133181.21695,             # end_time
-      {}                            # additional_info
-    ],
+    {
+      "config_id": 1,
+      "instance": null,
+      "seed": 209652396,
+      "budget": 2.7777777777777777,
+      "cost": 2147483647.0,
+      "time": 0.0,
+      "cpu_time": 0.0,
+      "status": 0,
+      "starttime": 0.0,
+      "endtime": 0.0,
+      "additional_info": {}
+    },
     ...
   ]
+
 ```
 
 `configs` is a human-readable dictionary of configurations, where the keys are the one-based `config_id`. It is important to note that in `runhistory.json`, the indexing is zero-based.

diff --git a/docs/advanced_usage/9_parallelism.md b/docs/advanced_usage/9_parallelism.md
@@ -23,7 +23,7 @@ SMAC supports multiple workers natively via Dask. Just specify ``n_workers`` in
 ## Running on a Cluster
 
 You can also pass a custom dask client, e.g. to run on a slurm cluster.
-See our [parallelism example](../examples/1%20Basics/7_parallelization_cluster.html).
+See our [parallelism example](../examples/1%20Basics/7_parallelization_cluster.md).
 
 !!! warning
 

diff --git a/docs/conf.py b/docs/conf.py
@@ -12,6 +12,7 @@
     "version": version,
     "versions": {
         f"v{version}": "#",
+        "v2.3.0": "https://automl.github.io/SMAC3/v2.3.0/",
         "v2.2.0": "https://automl.github.io/SMAC3/v2.2.0/",
         "v2.1.0": "https://automl.github.io/SMAC3/v2.1.0/",
         "v2.0.1": "https://automl.github.io/SMAC3/v2.0.1/",

diff --git a/examples/1_basics/1_quadratic_function.py b/examples/1_basics/1_quadratic_function.py
@@ -16,7 +16,7 @@
 from smac.facade.hyperparameter_optimization_facade import HyperparameterOptimizationFacade as HPOFacade
 from smac import RunHistory, Scenario
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/1_basics/2_svm_cv.py b/examples/1_basics/2_svm_cv.py
@@ -15,7 +15,7 @@
 
 from smac import HyperparameterOptimizationFacade, Scenario
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/1_basics/3_ask_and_tell.py b/examples/1_basics/3_ask_and_tell.py
@@ -9,7 +9,7 @@
 from smac import HyperparameterOptimizationFacade, Scenario
 from smac.runhistory.dataclasses import TrialValue
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/1_basics/4_callback.py b/examples/1_basics/4_callback.py
@@ -16,7 +16,7 @@
 from smac import Scenario
 from smac.runhistory import TrialInfo, TrialValue
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/1_basics/5_continue.py b/examples/1_basics/5_continue.py
@@ -23,7 +23,7 @@
 from smac.main.smbo import SMBO
 from smac.runhistory import TrialInfo, TrialValue
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/1_basics/6_priors.py b/examples/1_basics/6_priors.py
@@ -20,6 +20,7 @@
     ConfigurationSpace,
     NormalFloatHyperparameter,
     UniformIntegerHyperparameter,
+    UniformFloatHyperparameter,
 )
 from sklearn.datasets import load_digits
 from sklearn.exceptions import ConvergenceWarning
@@ -29,7 +30,7 @@
 from smac import HyperparameterOptimizationFacade, Scenario
 from smac.acquisition.function import PriorAcquisitionFunction
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 
@@ -38,7 +39,7 @@
 
 class MLP:
     @property
-    def configspace(self) -> ConfigurationSpace:
+    def prior_configspace(self) -> ConfigurationSpace:
         # Build Configuration Space which defines all parameters and their ranges.
         # To illustrate different parameter types,
         # we use continuous, integer and categorical parameters.
@@ -100,7 +101,67 @@ def configspace(self) -> ConfigurationSpace:
         )
 
         # Add all hyperparameters at once:
-        cs.add([n_layer, n_neurons, activation, optimizer, batch_size, learning_rate_init])
+        cs.add(
+            [n_layer, n_neurons, activation, optimizer, batch_size, learning_rate_init]
+        )
+
+        return cs
+
+    @property
+    def configspace(self) -> ConfigurationSpace:
+        # Build Configuration Space which defines all parameters and their ranges.
+        # To illustrate different parameter types,
+        # we use continuous, integer and categorical parameters.
+        cs = ConfigurationSpace()
+
+        # We do not have an educated belief on the number of layers beforehand
+        n_layer = UniformIntegerHyperparameter(
+            "n_layer",
+            lower=1,
+            upper=5,
+        )
+
+        # Define network width without a specific prior
+        n_neurons = UniformIntegerHyperparameter(
+            "n_neurons",
+            lower=8,
+            upper=256,
+        )
+
+        # Define activation functions without specific weights
+        activation = CategoricalHyperparameter(
+            "activation",
+            ["logistic", "tanh", "relu"],
+            default_value="relu",
+        )
+
+        # Define optimizer without specific weights
+        optimizer = CategoricalHyperparameter(
+            "optimizer",
+            ["sgd", "adam"],
+            default_value="adam",
+        )
+
+        # Define batch size without specific distribution
+        batch_size = UniformIntegerHyperparameter(
+            "batch_size",
+            16,
+            512,
+            default_value=128,
+        )
+
+        # Define learning rate range without log-normal prior
+        learning_rate_init = UniformFloatHyperparameter(
+            "learning_rate_init",
+            lower=1e-5,
+            upper=1.0,
+            default_value=1e-3,
+        )
+
+        # Add all hyperparameters at once:
+        cs.add(
+            [n_layer, n_neurons, activation, optimizer, batch_size, learning_rate_init]
+        )
 
         return cs
 
@@ -119,8 +180,12 @@ def train(self, config: Configuration, seed: int = 0) -> float:
             )
 
             # Returns the 5-fold cross validation accuracy
-            cv = StratifiedKFold(n_splits=5, random_state=seed, shuffle=True)  # to make CV splits consistent
-            score = cross_val_score(classifier, digits.data, digits.target, cv=cv, error_score="raise")
+            cv = StratifiedKFold(
+                n_splits=5, random_state=seed, shuffle=True
+            )  # to make CV splits consistent
+            score = cross_val_score(
+                classifier, digits.data, digits.target, cv=cv, error_score="raise"
+            )
 
         return 1 - np.mean(score)
 
@@ -140,7 +205,9 @@ def train(self, config: Configuration, seed: int = 0) -> float:
 
     # We define the prior acquisition function, which conduct the optimization using priors over the optimum
     acquisition_function = PriorAcquisitionFunction(
-        acquisition_function=HyperparameterOptimizationFacade.get_acquisition_function(scenario),
+        acquisition_function=HyperparameterOptimizationFacade.get_acquisition_function(
+            scenario
+        ),
         decay_beta=scenario.n_trials / 10,  # Proven solid value
     )
 

diff --git a/examples/1_basics/7_parallelization_cluster.py b/examples/1_basics/7_parallelization_cluster.py
@@ -32,7 +32,7 @@
 
 from smac import BlackBoxFacade, Scenario
 
-__copyright__ = "Copyright 2023, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/2_multi_fidelity/1_mlp_epochs.py b/examples/2_multi_fidelity/1_mlp_epochs.py
@@ -41,7 +41,7 @@
 from smac.intensifier.hyperband import Hyperband
 from smac.intensifier.successive_halving import SuccessiveHalving
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/2_multi_fidelity/2_sgd_datasets.py b/examples/2_multi_fidelity/2_sgd_datasets.py
@@ -27,7 +27,7 @@
 from smac import MultiFidelityFacade as MFFacade
 from smac import Scenario
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/2_multi_fidelity/3_specify_HB_via_total_budget.py b/examples/2_multi_fidelity/3_specify_HB_via_total_budget.py
@@ -20,7 +20,7 @@
 from smac import MultiFidelityFacade, RunHistory, Scenario
 from smac.intensifier.hyperband_utils import get_n_trials_for_hyperband_multifidelity
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/3_multi_objective/1_schaffer.py b/examples/3_multi_objective/1_schaffer.py
@@ -18,7 +18,7 @@
 from smac import Scenario
 from smac.facade import AbstractFacade
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
 
 

diff --git a/examples/3_multi_objective/2_parego.py b/examples/3_multi_objective/2_parego.py
@@ -33,7 +33,7 @@
 from smac.facade.abstract_facade import AbstractFacade
 from smac.multi_objective.parego import ParEGO
 
-__copyright__ = "Copyright 2021, AutoML.org Freiburg-Hannover"
+__copyright__ = "Copyright 2025, Leibniz University Hanover, Institute of AI"
 __license__ = "3-clause BSD"
Original file line number	Diff line number	Diff line change
Expand Up		@@ -19,4 +19,4 @@ The behavior can be controlled by setting the parameter ``overwrite`` in the fac
		and the old run is not affected.


		Please have a look at our [continue example](../examples/1%20Basics/5_continue.html).
		Please have a look at our [continue example](../examples/1%20Basics/5_continue.md).
Original file line number	Diff line number	Diff line change
Expand Up		@@ -16,4 +16,4 @@ and report the results of the trial.
		into the intensification process.


		Please have a look at our [ask-and-tell example](../examples/1%20Basics/3_ask_and_tell.html).
		Please have a look at our [ask-and-tell example](../examples/1%20Basics/3_ask_and_tell.md).