Merge pull request #77 from Erotemic/shields

Fixes for Travis

Author: fmfn

.coveragerc

@@ -0,0 +1,23 @@
+[run]
+branch = True
+source = bayes_opt
+
+[report]
+exclude_lines =
+    pragma: no cover
+    .*  # pragma: no cover
+    .*  # nocover
+    def __repr__
+    raise AssertionError
+    raise NotImplementedError
+    if 0:
+    verbose = .*
+    raise
+    pass
+    if __name__ == .__main__.:
+    print(.*)
+
+omit =
+    */setup.py
+    examples/*
+    tests/*

.gitignore

@@ -8,3 +8,14 @@ scratch/
 .DS_Store
 bo_eg*.png
 gif/
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*,cover
+.hypothesis/

.travis.yml

@@ -17,18 +17,15 @@ before_install:
   - pip install pytest-cov -U
   - pip install codecov -U
   #- pip install xdoctest -U
-#  - pip install delorean
 install:
   #- travis_retry python setup.py build develop
   - travis_retry pip install -e .
 script: 
-  - travis_wait pytest --cov-config .coveragerc --cov-report html --cov bayes_opt bayes_opt
-    #-p no:doctest --xdoctest --cov=ubelt ubelt
+  - travis_wait pytest --cov-config .coveragerc --cov-report html --cov=bayes_opt
+    #-p no:doctest --xdoctest
 after_success: 
   #- coveralls || echo "Coveralls upload failed"
   - codecov 
-#after_failure: 
-#  - cat failed_doctests.txt
 cache: 
     apt: true
     directories:

README.md

@@ -1,5 +1,9 @@
 # Bayesian Optimization
 
+[![Travis](https://img.shields.io/travis/fmfn/BayesianOptimization/master.svg?label=Travis%20CI)](https://travis-ci.org/fmfn/BayesianOptimization)
+[![Codecov](https://codecov.io/github/fmfn/BayesianOptimization/badge.svg?branch=master&service=github)](https://codecov.io/github/fmfn/BayesianOptimization?branch=master)
+[![Pypi](https://img.shields.io/pypi/v/bayesian-optimization.svg)](https://pypi.python.org/pypi/bayesian-optimization)
+
 Pure Python implementation of bayesian global optimization with gaussian
 processes.
 

bayes_opt/bayesian_optimization.py

@@ -64,6 +64,10 @@ def __init__(self, f, pbounds, random_state=None, verbose=1):
                            'max_params': None}
         self.res['all'] = {'values': [], 'params': []}
 
+        # non-public config for maximizing the aquisition function
+        # (used to speedup tests, but generally leave these as is)
+        self._acqkw = {'n_warmup': 100000, 'n_iter': 250}
+
         # Verbose
         self.verbose = verbose
 
@@ -251,7 +255,8 @@ def maximize(self,
                         gp=self.gp,
                         y_max=y_max,
                         bounds=self.space.bounds,
-                        random_state=self.random_state)
+                        random_state=self.random_state,
+                        **self._acqkw)
 
         # Print new header
         if self.verbose:
@@ -292,7 +297,8 @@ def maximize(self,
                             gp=self.gp,
                             y_max=y_max,
                             bounds=self.space.bounds,
-                            random_state=self.random_state)
+                            random_state=self.random_state,
+                            **self._acqkw)
 
             # Keep track of total number of iterations
             self.i += 1

bayes_opt/helpers.py

@@ -6,13 +6,13 @@
 from scipy.optimize import minimize
 
 
-def acq_max(ac, gp, y_max, bounds, random_state):
+def acq_max(ac, gp, y_max, bounds, random_state, n_warmup=100000, n_iter=250):
     """
     A function to find the maximum of the acquisition function
 
     It uses a combination of random sampling (cheap) and the 'L-BFGS-B'
-    optimization method. First by sampling 1e5 points at random, and then
-    running L-BFGS-B from 250 random starting points.
+    optimization method. First by sampling `n_warmup` (1e5) points at random,
+    and then running L-BFGS-B from `n_iter` (250) random starting points.
 
     Parameters
     ----------
@@ -28,6 +28,14 @@ def acq_max(ac, gp, y_max, bounds, random_state):
     :param bounds:
         The variables bounds to limit the search of the acq max.
 
+    :param random_state:
+        instance of np.RandomState random number generator
+
+    :param n_warmup:
+        number of times to randomly sample the aquisition function
+
+    :param n_iter:
+        number of times to run scipy.minimize
 
     Returns
     -------
@@ -36,14 +44,14 @@ def acq_max(ac, gp, y_max, bounds, random_state):
 
     # Warm up with random points
     x_tries = random_state.uniform(bounds[:, 0], bounds[:, 1],
-                                 size=(100000, bounds.shape[0]))
+                                   size=(n_warmup, bounds.shape[0]))
     ys = ac(x_tries, gp=gp, y_max=y_max)
     x_max = x_tries[ys.argmax()]
     max_acq = ys.max()
 
     # Explore the parameter space more throughly
     x_seeds = random_state.uniform(bounds[:, 0], bounds[:, 1],
-                                size=(250, bounds.shape[0]))
+                                   size=(n_iter, bounds.shape[0]))
     for x_try in x_seeds:
         # Find the minimum of minus the acquisition function
         res = minimize(lambda x: -ac(x.reshape(1, -1), gp=gp, y_max=y_max),

setup.py

@@ -8,7 +8,7 @@
     author='Fernando Nogueira',
     author_email="[email protected]",
     description='Bayesian Optimization package',
-    download_url = 'https://github.com/fmfn/BayesianOptimization/tarball/0.4',
+    download_url='https://github.com/fmfn/BayesianOptimization/tarball/0.4',
     install_requires=[
         "numpy >= 1.9.0",
         "scipy >= 0.14.0",

tests/test_bayesian_optimization.py

@@ -19,13 +19,17 @@ def test_bayes_opt_demo():
                               random_state=random_state,
                               verbose=0)
     gp_params = {'alpha': 1e-5, 'n_restarts_optimizer': 2}
-    bo.maximize(init_points=10, n_iter=10, acq='ucb', kappa=5, **gp_params)
+    # Change aquisition params to speedup optimization for testing purposes
+    bo._acqkw['n_iter'] = 5
+    bo._acqkw['n_warmup'] = 1000
+    bo.maximize(init_points=10, n_iter=5, acq='ucb', kappa=5, **gp_params)
     res = bo.space.max_point()
     max_params = res['max_params']
     max_val = res['max_val']
 
-    assert max_val > 1.2, 'function range is ~.2 - ~1.4, should be above 1.'
-    assert max_val / f.max() > .9, 'should be better than 90% of max val'
+    ratio = max_val / f.max()
+    assert max_val > 1.1, 'got {}, but should be > 1'.format(max_val)
+    assert ratio > .9, 'got {}, should be better than 90% of max val'.format(ratio)
 
     assert max_params['x'] > 300, 'should be in a peak area (around 300)'
     assert max_params['x'] < 400, 'should be in a peak area (around 300)'

tests/test_helper_functions.py

@@ -63,7 +63,7 @@ def test_acq_max_function_with_ucb_algo(self):
         self.setUp(kind='ucb', kappa=1.0, xi=1.0)
         max_arg = acq_max(
             self.util.utility, GP, self.y_max, bounds=np.array([[0, 1], [0, 1]]),
-            random_state=self.random_state
+            random_state=self.random_state, n_iter=20
         )
         _, brute_max_arg = brute_force_maximum(MESH, GP)
 
@@ -73,7 +73,7 @@ def test_ei_max_function_with_ucb_algo(self):
         self.setUp(kind='ei', kappa=1.0, xi=1e-6)
         max_arg = acq_max(
             self.util.utility, GP, self.y_max, bounds=np.array([[0, 1], [0, 1]]),
-            random_state=self.random_state
+            random_state=self.random_state, n_iter=20
         )
         _, brute_max_arg = brute_force_maximum(MESH, GP, kind='ei')