[FEATURE] Added copyright notice

RDUCB/README.md

@@ -2,6 +2,7 @@
 
 
 This repository accompanies a [ICML 2023 publication](https://arxiv.org/pdf/2301.12844.pdf) by Juliusz Ziomek and Haitham Bou-Ammar.
+The repository is largely based on code from [High-Dimensional Bayesian Optimization via Tree-Structured Additive Models ](https://github.com/eric-vader/HD-BO-Additive-Models), as such the code in this repository is released under the original MIT license (in the LICENSE file) giving copyright to Eric Han, except for the parts that have been added or substantially modified, which are released under MIT licence giving copyrights to Huawei Technologies Co., Ltd. Such parts are clearly marked in code by comments.
 
 ## Acknowledgements
 
@@ -35,9 +36,12 @@ Minimum System requirements:
 Prepare your environment:
 
 1. If you don't have it already, [Install Conda](https://docs.conda.io/projects/conda/en/latest/user-guide/install/linux.html)
-2. (Optional) Run `bash data/setup.sh` to download data from [NAS Benchmarks](https://github.com/automl/nas_benchmarks) and  into your home directory. You may skip this step if you are not running lpsolve and NAS-Bench-101 datasets.
-3. [Install MLflow](https://mlflow.org/) either on your system or in the base environment of Conda - `pip install mlflow`
-4. Build and test environment by using the command `mlflow run .` (this may take a while).
+2. [Install MLflow](https://mlflow.org/) either on your system or in the base environment of Conda - `pip install mlflow`
+3. Build and test environment by using the command `mlflow run .` (this may take a while).
+
+Optional steps:
+1. (Optional) Run `bash data/setup.sh` to download data from [NAS Benchmarks](https://github.com/automl/nas_benchmarks) and  into your home directory. You may skip this step if you are not running lpsolve and NAS-Bench-101 datasets.
+2. (Optional) Run `pip install -U pandas matplotlib seaborn` in your base environment for plotting the results.  You may skip this step if you do not need the plotting script.
 
 ## Running experiments from Paper
 
@@ -50,9 +54,9 @@ mlflow run . -P param_file=config/LassoBench/rducb.yml
 All the parameters of the run should be specified in the `.yml` file. However, to facilitate running batch jobs, one can include command line arguments to override the seed or subproblem. For example:
 
 ```
-mlflow run . -P param_file=config/LassoBench/rducb.yml -P seed=1 -P sub_benchmark="pick_data:leukemia"
+mlflow run . -P param_file=config/LassoBench/rducb.yml -P seed=1 -P sub_benchmark="pick_data:diabetes"
 ```
-will run the experiment as specified in `config/LassoBench/rducb.yml` but will overwrite the seed to 1 and pick_data to leukemia. See an example of workflow in `example.sh`
+will run the experiment as specified in `config/LassoBench/rducb.yml` but will overwrite the seed to 1 and pick_data to diabetes. See an example of workflow in `example.sh`
 
 
 ## Visualising the results

RDUCB/config/default.yml

@@ -14,18 +14,11 @@ algorithm_type:
     noise_var: 0.01
     param_n_iter: 16
 data_type:
-  Synthetic:
-    data_random_seed: 3
-    dimension: 15
-    fn_noise_var: 0.0225
-    graph_type: StarGraph
-    grid_params:
-      domain_lower: 0.0
-      domain_upper: 1.0
-      grid_size: 150
-    kernel_params:
-      lengthscale: 0.2
-      variance: 1.0
-
-n_iter: 1000
+  Hpolib:
+    aug_dimension: 14
+    data_random_seed: 2
+    fn_noise_var: 0
+    grid_size: 150
+    hpo_fn: Hartmann6Aug
+n_iter: 10
 n_rand: 10

RDUCB/example.sh

@@ -2,7 +2,7 @@
 mlflow run . -P param_file=config/examples/example1_random.yml -P seed=0
 mlflow run . -P param_file=config/examples/example1_random.yml -P seed=1
 
-# Run two seeds of RDUCBon example1 problem (Rosenbrock)
+# Run two seeds of RDUCB on example1 problem (Rosenbrock)
 mlflow run . -P param_file=config/examples/example1_rducb.yml -P seed=0
 mlflow run . -P param_file=config/examples/example1_rducb.yml -P seed=1
 

RDUCB/hdbo.yml

@@ -36,7 +36,7 @@ dependencies:
     - ConfigSpace==0.4.11 # this works with glibc 2.12 which xgpa is on
     - noisyopt
     - git+https://github.com/eric-vader/HPOlib2.git#egg=hpolib2
-    #- git+https://github.com/ksehic/LassoBench.git
+    - git+https://github.com/ksehic/LassoBench.git
     - tensorflow==2.2.0
 #    - ./tensorflow-2.2.0-cp38-cp38-linux_x86_64.whl
 #  - cudnn=7.6.5=cuda10.1_0

RDUCB/hdbo/algorithms.py

@@ -396,6 +396,30 @@ def FnOptimizer(self):
         return function_optimizer.Tree
 
 class RDUCB(BayesianOptimization, metaclass=Algorithm):
+    """
+    MIT License
+
+    Copyright (c) 2023 Huawei Technologies Co., Ltd
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
+
+    """
     __metaclass__ = Algorithm
     def __init__(self, **kwargs):
         self.random_graph = True

RDUCB/hdbo/datasets.py

@@ -709,6 +709,30 @@ def eval(self, x):
         return objective
 # =======================================================================
 class LassoBenchLoader(Loader, metaclass=LassoBenchlib):
+    """
+    MIT License
+
+    Copyright (c) 2023  Huawei Technologies Co., Ltd
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
+
+    """
     def __init__(self, pick_data, grid_size, fidelity, fixed_dims=0, **kwargs):
 
         self.f = ExecuteLassoBench(pick_data, grid_size, fidelity, fixed_dims)
@@ -719,6 +743,30 @@ def load(self):
 
 # Function to execute LP Solve
 class ExecuteLassoBench(Function):
+    """
+    MIT License
+
+    Copyright (c) 2023  Huawei Technologies Co., Ltd
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
+
+    """
     def __init__(self, pick_data, grid_size, fidelity=0, fixed_dims=0):
         import LassoBench
 
@@ -774,6 +822,30 @@ def eval(self, X):
         return self.hpo_fn(x)
 
 class Adversarial(Loader, metaclass=Simple):
+    """
+    MIT License
+
+    Copyright (c) 2023  Huawei Technologies Co., Ltd
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
+
+    """
     def __init__(self, grid_size, **kwargs):
         Loader.__init__(self, "Adversarial", **kwargs)
         logging.info("Using Adversarial function")
@@ -783,6 +855,30 @@ def load(self):
         return AdversairalFunction(self.domain), self.soln
 
 class AdversairalFunction(Function):
+    """
+    MIT License
+
+    Copyright (c) 2023  Huawei Technologies Co., Ltd
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
+
+    """
     def __init__(self, domain):
         Function.__init__(self, domain)
         self.graph = None

RDUCB/hdbo/graph_utils.py

@@ -77,6 +77,30 @@ def debug_graph(G):
 
 
 def get_random_graph(size, connection_draws=1):
+    """
+    MIT License
+
+    Copyright (c) 2023  Huawei Technologies Co., Ltd
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
+
+    """
 
     graph = nx.empty_graph(size)
     disjoint_set = DisjointSet()
@@ -95,24 +119,4 @@ def get_random_graph(size, connection_draws=1):
     return graph
 
 def sigmoid(x):
-    return 1 / (1 + math.exp(-x))
-
-
-if __name__ == "__main__":
-    import matplotlib.pyplot as plt
-    for i in range(100):
-        test_graph = get_random_graph(20, 5)
-        nx.draw_networkx(test_graph)
-        try:
-            nx.find_cycle(test_graph)
-            has_cycles = True
-        except nx.exception.NetworkXNoCycle:
-            has_cycles = False
-        print(i, test_graph.edges())
-        if has_cycles:
-            pass
-        else:
-            print('No cycles')
-
-        plt.savefig(f'test_graph{i}')
-        plt.clf()
+    return 1 / (1 + math.exp(-x))

RDUCB/plot.py

@@ -1,3 +1,28 @@
+"""
+    MIT License
+
+    Copyright (c) 2023  Huawei Technologies Co., Ltd
+
+    Permission is hereby granted, free of charge, to any person obtaining a copy
+    of this software and associated documentation files (the "Software"), to deal
+    in the Software without restriction, including without limitation the rights
+    to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+    copies of the Software, and to permit persons to whom the Software is
+    furnished to do so, subject to the following conditions:
+
+    The above copyright notice and this permission notice shall be included in all
+    copies or substantial portions of the Software.
+
+    THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+    IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+    FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+    AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+    LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+    OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+    SOFTWARE.
+
+"""
+
 import pandas as pd
 import seaborn as sns
 import os
@@ -76,7 +101,7 @@ def get_matching_runs(data_type, data_name, exehashs, ignore_error=False, run_le
                 if type(data) == type(None):
                     data = new_data
                 else:
-                    data = data.append(new_data, ignore_index=True)
+                    data = pd.concat([data, new_data], ignore_index=True)
 
         print(f'Found {run_no} runs for {name}')