ANL-CEEESA/MIPLearn
1
0
Fork 0
mirror of https://github.com/ANL-CEEESA/MIPLearn.git synced 2025-12-06 01:18:52 -06:00
Code Issues Packages Projects Releases Wiki Activity

Reformat source code with Black; add pre-commit hooks and CI checks

Browse Source
This commit is contained in:
Alinson S. Xavier
2020-12-05 10:59:33 -06:00
parent 3823931382
commit d99600f101
49 changed files with 1291 additions and 972 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
.github/workflows/lint.yml vendored Normal file
View File

@@ -0,0 +1,11 @@
name: Lint
on: [push, pull_request]
jobs:
lint:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v2
- uses: actions/setup-python@v2
- uses: psf/black@stable

6
.pre-commit-config.yaml Normal file
View File

@@ -0,0 +1,6 @@
repos:
- repo: https://github.com/ambv/black
rev: stable
hooks:
- id: black
args: ["--check"]

3
Makefile
View File

@@ -34,6 +34,9 @@ install:
uninstall: uninstall:
$(PIP) uninstall miplearn $(PIP) uninstall miplearn
reformat:
$(PYTHON) -m black miplearn
test: test:
$(PYTEST) $(PYTEST_ARGS) $(PYTEST) $(PYTEST_ARGS)

10
miplearn/__init__.py
View File

@@ -2,10 +2,12 @@
# Copyright (C) 2020, UChicago Argonne, LLC. All rights reserved. # Copyright (C) 2020, UChicago Argonne, LLC. All rights reserved.
# Released under the modified BSD license. See COPYING.md for more details. # Released under the modified BSD license. See COPYING.md for more details.
from .extractors import (SolutionExtractor, from .extractors import (
InstanceFeaturesExtractor, SolutionExtractor,
ObjectiveValueExtractor, InstanceFeaturesExtractor,
VariableFeaturesExtractor) ObjectiveValueExtractor,
VariableFeaturesExtractor,
)
from .components.component import Component from .components.component import Component
from .components.objective import ObjectiveValueComponent from .components.objective import ObjectiveValueComponent

238
miplearn/benchmark.py
View File

@@ -19,43 +19,53 @@ class BenchmarkRunner:
assert isinstance(solver, LearningSolver) assert isinstance(solver, LearningSolver)
self.solvers = solvers self.solvers = solvers
self.results = None self.results = None
def solve(self, instances, tee=False): def solve(self, instances, tee=False):
for (solver_name, solver) in self.solvers.items(): for (solver_name, solver) in self.solvers.items():
for i in tqdm(range(len((instances)))): for i in tqdm(range(len((instances)))):
results = solver.solve(deepcopy(instances[i]), tee=tee) results = solver.solve(deepcopy(instances[i]), tee=tee)
self._push_result(results, solver=solver, solver_name=solver_name, instance=i) self._push_result(
results,
solver=solver,
solver_name=solver_name,
instance=i,
)
def parallel_solve(self, def parallel_solve(
instances, self,
n_jobs=1, instances,
n_trials=1, n_jobs=1,
index_offset=0, n_trials=1,
): index_offset=0,
):
self._silence_miplearn_logger() self._silence_miplearn_logger()
trials = instances * n_trials trials = instances * n_trials
for (solver_name, solver) in self.solvers.items(): for (solver_name, solver) in self.solvers.items():
results = solver.parallel_solve(trials, results = solver.parallel_solve(
n_jobs=n_jobs, trials,
label="Solve (%s)" % solver_name, n_jobs=n_jobs,
output=None) label="Solve (%s)" % solver_name,
output=None,
)
for i in range(len(trials)): for i in range(len(trials)):
idx = (i % len(instances)) + index_offset idx = (i % len(instances)) + index_offset
self._push_result(results[i], self._push_result(
solver=solver, results[i],
solver_name=solver_name, solver=solver,
instance=idx) solver_name=solver_name,
instance=idx,
)
self._restore_miplearn_logger() self._restore_miplearn_logger()
def raw_results(self): def raw_results(self):
return self.results return self.results
def save_results(self, filename): def save_results(self, filename):
self.results.to_csv(filename) self.results.to_csv(filename)
def load_results(self, filename): def load_results(self, filename):
self.results = pd.read_csv(filename, index_col=0) self.results = pd.read_csv(filename, index_col=0)
def load_state(self, filename): def load_state(self, filename):
for (solver_name, solver) in self.solvers.items(): for (solver_name, solver) in self.solvers.items():
solver.load_state(filename) solver.load_state(filename)
@@ -63,62 +73,69 @@ class BenchmarkRunner:
def fit(self, training_instances): def fit(self, training_instances):
for (solver_name, solver) in self.solvers.items(): for (solver_name, solver) in self.solvers.items():
solver.fit(training_instances) solver.fit(training_instances)
def _push_result(self, result, solver, solver_name, instance): def _push_result(self, result, solver, solver_name, instance):
if self.results is None: if self.results is None:
self.results = pd.DataFrame(columns=["Solver", self.results = pd.DataFrame(
"Instance", columns=[
"Wallclock Time", "Solver",
"Lower Bound", "Instance",
"Upper Bound", "Wallclock Time",
"Gap", "Lower Bound",
"Nodes", "Upper Bound",
"Mode", "Gap",
"Sense", "Nodes",
"Predicted LB", "Mode",
"Predicted UB", "Sense",
]) "Predicted LB",
"Predicted UB",
]
)
lb = result["Lower bound"] lb = result["Lower bound"]
ub = result["Upper bound"] ub = result["Upper bound"]
gap = (ub - lb) / lb gap = (ub - lb) / lb
if "Predicted LB" not in result: if "Predicted LB" not in result:
result["Predicted LB"] = float("nan") result["Predicted LB"] = float("nan")
result["Predicted UB"] = float("nan") result["Predicted UB"] = float("nan")
self.results = self.results.append({ self.results = self.results.append(
"Solver": solver_name, {
"Instance": instance, "Solver": solver_name,
"Wallclock Time": result["Wallclock time"], "Instance": instance,
"Lower Bound": lb, "Wallclock Time": result["Wallclock time"],
"Upper Bound": ub, "Lower Bound": lb,
"Gap": gap, "Upper Bound": ub,
"Nodes": result["Nodes"], "Gap": gap,
"Mode": solver.mode, "Nodes": result["Nodes"],
"Sense": result["Sense"], "Mode": solver.mode,
"Predicted LB": result["Predicted LB"], "Sense": result["Sense"],
"Predicted UB": result["Predicted UB"], "Predicted LB": result["Predicted LB"],
}, ignore_index=True) "Predicted UB": result["Predicted UB"],
},
ignore_index=True,
)
groups = self.results.groupby("Instance") groups = self.results.groupby("Instance")
best_lower_bound = groups["Lower Bound"].transform("max") best_lower_bound = groups["Lower Bound"].transform("max")
best_upper_bound = groups["Upper Bound"].transform("min") best_upper_bound = groups["Upper Bound"].transform("min")
best_gap = groups["Gap"].transform("min") best_gap = groups["Gap"].transform("min")
best_nodes = np.maximum(1, groups["Nodes"].transform("min")) best_nodes = np.maximum(1, groups["Nodes"].transform("min"))
best_wallclock_time = groups["Wallclock Time"].transform("min") best_wallclock_time = groups["Wallclock Time"].transform("min")
self.results["Relative Lower Bound"] = \ self.results["Relative Lower Bound"] = (
self.results["Lower Bound"] / best_lower_bound self.results["Lower Bound"] / best_lower_bound
self.results["Relative Upper Bound"] = \ )
self.results["Upper Bound"] / best_upper_bound self.results["Relative Upper Bound"] = (
self.results["Relative Wallclock Time"] = \ self.results["Upper Bound"] / best_upper_bound
self.results["Wallclock Time"] / best_wallclock_time )
self.results["Relative Gap"] = \ self.results["Relative Wallclock Time"] = (
self.results["Gap"] / best_gap self.results["Wallclock Time"] / best_wallclock_time
self.results["Relative Nodes"] = \ )
self.results["Nodes"] / best_nodes self.results["Relative Gap"] = self.results["Gap"] / best_gap
self.results["Relative Nodes"] = self.results["Nodes"] / best_nodes
def save_chart(self, filename): def save_chart(self, filename):
import matplotlib.pyplot as plt import matplotlib.pyplot as plt
import seaborn as sns import seaborn as sns
from numpy import median from numpy import median
sns.set_style("whitegrid") sns.set_style("whitegrid")
sns.set_palette("Blues_r") sns.set_palette("Blues_r")
results = self.raw_results() results = self.raw_results()
@@ -134,71 +151,76 @@ class BenchmarkRunner:
obj_column = "Lower Bound" obj_column = "Lower Bound"
predicted_obj_column = "Predicted LB" predicted_obj_column = "Predicted LB"
fig, (ax1, ax2, ax3, ax4) = plt.subplots(nrows=1, fig, (ax1, ax2, ax3, ax4) = plt.subplots(
ncols=4, nrows=1,
figsize=(12,4), ncols=4,
gridspec_kw={'width_ratios': [2, 1, 1, 2]}) figsize=(12, 4),
gridspec_kw={"width_ratios": [2, 1, 1, 2]},
)
# Figure 1: Solver x Wallclock Time # Figure 1: Solver x Wallclock Time
sns.stripplot(x="Solver", sns.stripplot(
y="Wallclock Time", x="Solver",
data=results, y="Wallclock Time",
ax=ax1, data=results,
jitter=0.25, ax=ax1,
size=4.0, jitter=0.25,
) size=4.0,
sns.barplot(x="Solver", )
y="Wallclock Time", sns.barplot(
data=results, x="Solver",
ax=ax1, y="Wallclock Time",
errwidth=0., data=results,
alpha=0.4, ax=ax1,
estimator=median, errwidth=0.0,
) alpha=0.4,
ax1.set(ylabel='Wallclock Time (s)') estimator=median,
)
ax1.set(ylabel="Wallclock Time (s)")
# Figure 2: Solver x Gap (%) # Figure 2: Solver x Gap (%)
ax2.set_ylim(-0.5, 5.5) ax2.set_ylim(-0.5, 5.5)
sns.stripplot(x="Solver", sns.stripplot(
y="Gap (%)", x="Solver",
jitter=0.25, y="Gap (%)",
data=results[results["Mode"] != "heuristic"], jitter=0.25,
ax=ax2, data=results[results["Mode"] != "heuristic"],
size=4.0, ax=ax2,
) size=4.0,
)
# Figure 3: Solver x Primal Value # Figure 3: Solver x Primal Value
ax3.set_ylim(0.95,1.05) ax3.set_ylim(0.95, 1.05)
sns.stripplot(x="Solver", sns.stripplot(
y=primal_column, x="Solver",
jitter=0.25, y=primal_column,
data=results[results["Mode"] == "heuristic"], jitter=0.25,
ax=ax3, data=results[results["Mode"] == "heuristic"],
) ax=ax3,
)
# Figure 4: Predicted vs Actual Objective Value # Figure 4: Predicted vs Actual Objective Value
sns.scatterplot(x=obj_column, sns.scatterplot(
y=predicted_obj_column, x=obj_column,
hue="Solver", y=predicted_obj_column,
data=results[results["Mode"] != "heuristic"], hue="Solver",
ax=ax4, data=results[results["Mode"] != "heuristic"],
) ax=ax4,
)
xlim, ylim = ax4.get_xlim(), ax4.get_ylim() xlim, ylim = ax4.get_xlim(), ax4.get_ylim()
ax4.plot([-1e10, 1e10], [-1e10, 1e10], ls='-', color="#cccccc") ax4.plot([-1e10, 1e10], [-1e10, 1e10], ls="-", color="#cccccc")
ax4.set_xlim(xlim) ax4.set_xlim(xlim)
ax4.set_ylim(ylim) ax4.set_ylim(ylim)
ax4.get_legend().remove() ax4.get_legend().remove()
fig.tight_layout() fig.tight_layout()
plt.savefig(filename, bbox_inches='tight', dpi=150) plt.savefig(filename, bbox_inches="tight", dpi=150)
def _silence_miplearn_logger(self): def _silence_miplearn_logger(self):
miplearn_logger = logging.getLogger("miplearn") miplearn_logger = logging.getLogger("miplearn")
self.prev_log_level = miplearn_logger.getEffectiveLevel() self.prev_log_level = miplearn_logger.getEffectiveLevel()
miplearn_logger.setLevel(logging.WARNING) miplearn_logger.setLevel(logging.WARNING)
def _restore_miplearn_logger(self): def _restore_miplearn_logger(self):
miplearn_logger = logging.getLogger("miplearn") miplearn_logger = logging.getLogger("miplearn")
miplearn_logger.setLevel(self.prev_log_level) miplearn_logger.setLevel(self.prev_log_level)

13
miplearn/classifiers/adaptive.py
View File

@@ -22,9 +22,11 @@ class AdaptiveClassifier(Classifier):
based on its cross-validation score on a particular training data set. based on its cross-validation score on a particular training data set.
""" """
def __init__(self, def __init__(
candidates=None, self,
evaluator=ClassifierEvaluator()): candidates=None,
evaluator=ClassifierEvaluator(),
):
""" """
Initializes the meta-classifier. Initializes the meta-classifier.
""" """
@@ -35,14 +37,13 @@ class AdaptiveClassifier(Classifier):
"min samples": 100, "min samples": 100,
}, },
"logistic": { "logistic": {
"classifier": make_pipeline(StandardScaler(), "classifier": make_pipeline(StandardScaler(), LogisticRegression()),
LogisticRegression()),
"min samples": 30, "min samples": 30,
}, },
"counting": { "counting": {
"classifier": CountingClassifier(), "classifier": CountingClassifier(),
"min samples": 0, "min samples": 0,
} },
} }
self.candidates = candidates self.candidates = candidates
self.evaluator = evaluator self.evaluator = evaluator

3
miplearn/classifiers/counting.py
View File

@@ -21,8 +21,7 @@ class CountingClassifier(Classifier):
self.mean = np.mean(y_train) self.mean = np.mean(y_train)
def predict_proba(self, x_test): def predict_proba(self, x_test):
return np.array([[1 - self.mean, self.mean] return np.array([[1 - self.mean, self.mean] for _ in range(x_test.shape[0])])
for _ in range(x_test.shape[0])])
def __repr__(self): def __repr__(self):
return "CountingClassifier(mean=%s)" % self.mean return "CountingClassifier(mean=%s)" % self.mean

51
miplearn/classifiers/cv.py
View File

@@ -11,6 +11,7 @@ from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import cross_val_score from sklearn.model_selection import cross_val_score
import logging import logging
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
@@ -28,12 +29,14 @@ class CrossValidatedClassifier(Classifier):
acceptable. Other numbers are a linear interpolation of these two extremes. acceptable. Other numbers are a linear interpolation of these two extremes.
""" """
def __init__(self, def __init__(
classifier=LogisticRegression(), self,
threshold=0.75, classifier=LogisticRegression(),
constant=0.0, threshold=0.75,
cv=5, constant=0.0,
scoring='accuracy'): cv=5,
scoring="accuracy",
):
self.classifier = None self.classifier = None
self.classifier_prototype = classifier self.classifier_prototype = classifier
self.constant = constant self.constant = constant
@@ -45,24 +48,36 @@ class CrossValidatedClassifier(Classifier):
# Calculate dummy score and absolute score threshold # Calculate dummy score and absolute score threshold
y_train_avg = np.average(y_train) y_train_avg = np.average(y_train)
dummy_score = max(y_train_avg, 1 - y_train_avg) dummy_score = max(y_train_avg, 1 - y_train_avg)
absolute_threshold = 1. * self.threshold + dummy_score * (1 - self.threshold) absolute_threshold = 1.0 * self.threshold + dummy_score * (1 - self.threshold)
# Calculate cross validation score and decide which classifier to use # Calculate cross validation score and decide which classifier to use
clf = deepcopy(self.classifier_prototype) clf = deepcopy(self.classifier_prototype)
cv_score = float(np.mean(cross_val_score(clf, cv_score = float(
x_train, np.mean(
y_train, cross_val_score(
cv=self.cv, clf,
scoring=self.scoring))) x_train,
y_train,
cv=self.cv,
scoring=self.scoring,
)
)
)
if cv_score >= absolute_threshold: if cv_score >= absolute_threshold:
logger.debug("cv_score is above threshold (%.2f >= %.2f); keeping" % logger.debug(
(cv_score, absolute_threshold)) "cv_score is above threshold (%.2f >= %.2f); keeping"
% (cv_score, absolute_threshold)
)
self.classifier = clf self.classifier = clf
else: else:
logger.debug("cv_score is below threshold (%.2f < %.2f); discarding" % logger.debug(
(cv_score, absolute_threshold)) "cv_score is below threshold (%.2f < %.2f); discarding"
self.classifier = DummyClassifier(strategy="constant", % (cv_score, absolute_threshold)
constant=self.constant) )
self.classifier = DummyClassifier(
strategy="constant",
constant=self.constant,
)
# Train chosen classifier # Train chosen classifier
self.classifier.fit(x_train, y_train) self.classifier.fit(x_train, y_train)

3
miplearn/classifiers/tests/test_counting.py
View File

@@ -12,7 +12,6 @@ E = 0.1
def test_counting(): def test_counting():
clf = CountingClassifier() clf = CountingClassifier()
clf.fit(np.zeros((8, 25)), [0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0]) clf.fit(np.zeros((8, 25)), [0.0, 0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 1.0])
expected_proba = np.array([[0.375, 0.625], expected_proba = np.array([[0.375, 0.625], [0.375, 0.625]])
[0.375, 0.625]])
actual_proba = clf.predict_proba(np.zeros((2, 25))) actual_proba = clf.predict_proba(np.zeros((2, 25)))
assert norm(actual_proba - expected_proba) < E assert norm(actual_proba - expected_proba) < E

36
miplearn/classifiers/tests/test_cv.py
View File

@@ -13,34 +13,36 @@ E = 0.1
def test_cv(): def test_cv():
# Training set: label is true if point is inside a 2D circle # Training set: label is true if point is inside a 2D circle
x_train = np.array([[x1, x2] x_train = np.array([[x1, x2] for x1 in range(-10, 11) for x2 in range(-10, 11)])
for x1 in range(-10, 11)
for x2 in range(-10, 11)])
x_train = StandardScaler().fit_transform(x_train) x_train = StandardScaler().fit_transform(x_train)
n_samples = x_train.shape[0] n_samples = x_train.shape[0]
y_train = np.array([1.0 if x1*x1 + x2*x2 <= 100 else 0.0 y_train = np.array(
for x1 in range(-10, 11) [
for x2 in range(-10, 11)]) 1.0 if x1 * x1 + x2 * x2 <= 100 else 0.0
for x1 in range(-10, 11)
for x2 in range(-10, 11)
]
)
# Support vector machines with linear kernels do not perform well on this # Support vector machines with linear kernels do not perform well on this
# data set, so predictor should return the given constant. # data set, so predictor should return the given constant.
clf = CrossValidatedClassifier(classifier=SVC(probability=True, clf = CrossValidatedClassifier(
random_state=42), classifier=SVC(probability=True, random_state=42),
threshold=0.90, threshold=0.90,
constant=0.0, constant=0.0,
cv=30) cv=30,
)
clf.fit(x_train, y_train) clf.fit(x_train, y_train)
assert norm(np.zeros(n_samples) - clf.predict(x_train)) < E assert norm(np.zeros(n_samples) - clf.predict(x_train)) < E
# Support vector machines with quadratic kernels perform almost perfectly # Support vector machines with quadratic kernels perform almost perfectly
# on this data set, so predictor should return their prediction. # on this data set, so predictor should return their prediction.
clf = CrossValidatedClassifier(classifier=SVC(probability=True, clf = CrossValidatedClassifier(
kernel='poly', classifier=SVC(probability=True, kernel="poly", degree=2, random_state=42),
degree=2, threshold=0.90,
random_state=42), cv=30,
threshold=0.90, )
cv=30)
clf.fit(x_train, y_train) clf.fit(x_train, y_train)
print(y_train - clf.predict(x_train)) print(y_train - clf.predict(x_train))
assert norm(y_train - clf.predict(x_train)) < E assert norm(y_train - clf.predict(x_train)) < E

1
miplearn/classifiers/tests/test_evaluator.py
View File

@@ -17,4 +17,3 @@ def test_evaluator():
ev = ClassifierEvaluator() ev = ClassifierEvaluator()
assert ev.evaluate(clf_a, x_train, y_train) == 1.0 assert ev.evaluate(clf_a, x_train, y_train) == 1.0
assert ev.evaluate(clf_b, x_train, y_train) == 0.5 assert ev.evaluate(clf_b, x_train, y_train) == 0.5

17
miplearn/classifiers/tests/test_threshold.py
View File

@@ -11,12 +11,16 @@ from miplearn.classifiers.threshold import MinPrecisionThreshold
def test_threshold_dynamic(): def test_threshold_dynamic():
clf = Mock(spec=Classifier) clf = Mock(spec=Classifier)
clf.predict_proba = Mock(return_value=np.array([ clf.predict_proba = Mock(
[0.10, 0.90], return_value=np.array(
[0.10, 0.90], [
[0.20, 0.80], [0.10, 0.90],
[0.30, 0.70], [0.10, 0.90],
])) [0.20, 0.80],
[0.30, 0.70],
]
)
)
x_train = np.array([0, 1, 2, 3]) x_train = np.array([0, 1, 2, 3])
y_train = np.array([1, 1, 0, 0]) y_train = np.array([1, 1, 0, 0])
@@ -31,4 +35,3 @@ def test_threshold_dynamic():
threshold = MinPrecisionThreshold(min_precision=0.00) threshold = MinPrecisionThreshold(min_precision=0.00)
assert threshold.find(clf, x_train, y_train) == 0.70 assert threshold.find(clf, x_train, y_train) == 0.70

14
miplearn/classifiers/threshold.py
View File

@@ -30,11 +30,15 @@ class MinPrecisionThreshold(DynamicThreshold):
def find(self, clf, x_train, y_train): def find(self, clf, x_train, y_train):
proba = clf.predict_proba(x_train) proba = clf.predict_proba(x_train)
assert isinstance(proba, np.ndarray), \ assert isinstance(proba, np.ndarray), "classifier should return numpy array"
"classifier should return numpy array" assert proba.shape == (
assert proba.shape == (x_train.shape[0], 2), \ x_train.shape[0],
"classifier should return (%d,%d)-shaped array, not %s" % ( 2,
x_train.shape[0], 2, str(proba.shape)) ), "classifier should return (%d,%d)-shaped array, not %s" % (
x_train.shape[0],
2,
str(proba.shape),
)
fps, tps, thresholds = _binary_clf_curve(y_train, proba[:, 1]) fps, tps, thresholds = _binary_clf_curve(y_train, proba[:, 1])
precision = tps / (tps + fps) precision = tps / (tps + fps)

6
miplearn/components/component.py
View File

@@ -9,15 +9,15 @@ class Component(ABC):
""" """
A Component is an object which adds functionality to a LearningSolver. A Component is an object which adds functionality to a LearningSolver.
""" """
@abstractmethod @abstractmethod
def before_solve(self, solver, instance, model): def before_solve(self, solver, instance, model):
pass pass
@abstractmethod @abstractmethod
def after_solve(self, solver, instance, model, results): def after_solve(self, solver, instance, model, results):
pass pass
@abstractmethod @abstractmethod
def fit(self, training_instances): def fit(self, training_instances):
pass pass

30
miplearn/components/cuts.py
View File

@@ -18,10 +18,12 @@ class UserCutsComponent(Component):
""" """
A component that predicts which user cuts to enforce. A component that predicts which user cuts to enforce.
""" """
def __init__(self, def __init__(
classifier=CountingClassifier(), self,
threshold=0.05): classifier=CountingClassifier(),
threshold=0.05,
):
self.violations = set() self.violations = set()
self.count = {} self.count = {}
self.n_samples = 0 self.n_samples = 0
@@ -40,7 +42,7 @@ class UserCutsComponent(Component):
def after_solve(self, solver, instance, model, results): def after_solve(self, solver, instance, model, results):
pass pass
def fit(self, training_instances): def fit(self, training_instances):
logger.debug("Fitting...") logger.debug("Fitting...")
features = InstanceFeaturesExtractor().extract(training_instances) features = InstanceFeaturesExtractor().extract(training_instances)
@@ -56,10 +58,11 @@ class UserCutsComponent(Component):
violation_to_instance_idx[v] = [] violation_to_instance_idx[v] = []
violation_to_instance_idx[v] += [idx] violation_to_instance_idx[v] += [idx]
for (v, classifier) in tqdm(self.classifiers.items(), for (v, classifier) in tqdm(
desc="Fit (user cuts)", self.classifiers.items(),
disable=not sys.stdout.isatty(), desc="Fit (user cuts)",
): disable=not sys.stdout.isatty(),
):
logger.debug("Training: %s" % (str(v))) logger.debug("Training: %s" % (str(v)))
label = np.zeros(len(training_instances)) label = np.zeros(len(training_instances))
label[violation_to_instance_idx[v]] = 1.0 label[violation_to_instance_idx[v]] = 1.0
@@ -79,10 +82,11 @@ class UserCutsComponent(Component):
all_violations = set() all_violations = set()
for instance in instances: for instance in instances:
all_violations |= set(instance.found_violated_user_cuts) all_violations |= set(instance.found_violated_user_cuts)
for idx in tqdm(range(len(instances)), for idx in tqdm(
desc="Evaluate (lazy)", range(len(instances)),
disable=not sys.stdout.isatty(), desc="Evaluate (lazy)",
): disable=not sys.stdout.isatty(),
):
instance = instances[idx] instance = instances[idx]
condition_positive = set(instance.found_violated_user_cuts) condition_positive = set(instance.found_violated_user_cuts)
condition_negative = all_violations - condition_positive condition_negative = all_violations - condition_positive

30
miplearn/components/lazy_dynamic.py
View File

@@ -18,10 +18,12 @@ class DynamicLazyConstraintsComponent(Component):
""" """
A component that predicts which lazy constraints to enforce. A component that predicts which lazy constraints to enforce.
""" """
def __init__(self, def __init__(
classifier=CountingClassifier(), self,
threshold=0.05): classifier=CountingClassifier(),
threshold=0.05,
):
self.violations = set() self.violations = set()
self.count = {} self.count = {}
self.n_samples = 0 self.n_samples = 0
@@ -52,7 +54,7 @@ class DynamicLazyConstraintsComponent(Component):
def after_solve(self, solver, instance, model, results): def after_solve(self, solver, instance, model, results):
pass pass
def fit(self, training_instances): def fit(self, training_instances):
logger.debug("Fitting...") logger.debug("Fitting...")
features = InstanceFeaturesExtractor().extract(training_instances) features = InstanceFeaturesExtractor().extract(training_instances)
@@ -68,10 +70,11 @@ class DynamicLazyConstraintsComponent(Component):
violation_to_instance_idx[v] = [] violation_to_instance_idx[v] = []
violation_to_instance_idx[v] += [idx] violation_to_instance_idx[v] += [idx]
for (v, classifier) in tqdm(self.classifiers.items(), for (v, classifier) in tqdm(
desc="Fit (lazy)", self.classifiers.items(),
disable=not sys.stdout.isatty(), desc="Fit (lazy)",
): disable=not sys.stdout.isatty(),
):
logger.debug("Training: %s" % (str(v))) logger.debug("Training: %s" % (str(v)))
label = np.zeros(len(training_instances)) label = np.zeros(len(training_instances))
label[violation_to_instance_idx[v]] = 1.0 label[violation_to_instance_idx[v]] = 1.0
@@ -91,10 +94,11 @@ class DynamicLazyConstraintsComponent(Component):
all_violations = set() all_violations = set()
for instance in instances: for instance in instances:
all_violations |= set(instance.found_violated_lazy_constraints) all_violations |= set(instance.found_violated_lazy_constraints)
for idx in tqdm(range(len(instances)), for idx in tqdm(
desc="Evaluate (lazy)", range(len(instances)),
disable=not sys.stdout.isatty(), desc="Evaluate (lazy)",
): disable=not sys.stdout.isatty(),
):
instance = instances[idx] instance = instances[idx]
condition_positive = set(instance.found_violated_lazy_constraints) condition_positive = set(instance.found_violated_lazy_constraints)
condition_negative = all_violations - condition_positive condition_negative = all_violations - condition_positive

53
miplearn/components/lazy_static.py
View File

@@ -19,13 +19,14 @@ class LazyConstraint:
class StaticLazyConstraintsComponent(Component): class StaticLazyConstraintsComponent(Component):
def __init__(self, def __init__(
classifier=CountingClassifier(), self,
threshold=0.05, classifier=CountingClassifier(),
use_two_phase_gap=True, threshold=0.05,
large_gap=1e-2, use_two_phase_gap=True,
violation_tolerance=-0.5, large_gap=1e-2,
): violation_tolerance=-0.5,
):
self.threshold = threshold self.threshold = threshold
self.classifier_prototype = classifier self.classifier_prototype = classifier
self.classifiers = {} self.classifiers = {}
@@ -74,32 +75,38 @@ class StaticLazyConstraintsComponent(Component):
logger.debug("Finding violated lazy constraints...") logger.debug("Finding violated lazy constraints...")
constraints_to_add = [] constraints_to_add = []
for c in self.pool: for c in self.pool:
if not solver.internal_solver.is_constraint_satisfied(c.obj, if not solver.internal_solver.is_constraint_satisfied(
tol=self.violation_tolerance): c.obj, tol=self.violation_tolerance
):
constraints_to_add.append(c) constraints_to_add.append(c)
for c in constraints_to_add: for c in constraints_to_add:
self.pool.remove(c) self.pool.remove(c)
solver.internal_solver.add_constraint(c.obj) solver.internal_solver.add_constraint(c.obj)
instance.found_violated_lazy_constraints += [c.cid] instance.found_violated_lazy_constraints += [c.cid]
if len(constraints_to_add) > 0: if len(constraints_to_add) > 0:
logger.info("%8d lazy constraints added %8d in the pool" % (len(constraints_to_add), len(self.pool))) logger.info(
"%8d lazy constraints added %8d in the pool"
% (len(constraints_to_add), len(self.pool))
)
return True return True
else: else:
return False return False
def fit(self, training_instances): def fit(self, training_instances):
training_instances = [t training_instances = [
for t in training_instances t
if hasattr(t, "found_violated_lazy_constraints")] for t in training_instances
if hasattr(t, "found_violated_lazy_constraints")
]
logger.debug("Extracting x and y...") logger.debug("Extracting x and y...")
x = self.x(training_instances) x = self.x(training_instances)
y = self.y(training_instances) y = self.y(training_instances)
logger.debug("Fitting...") logger.debug("Fitting...")
for category in tqdm(x.keys(), for category in tqdm(
desc="Fit (lazy)", x.keys(), desc="Fit (lazy)", disable=not sys.stdout.isatty()
disable=not sys.stdout.isatty()): ):
if category not in self.classifiers: if category not in self.classifiers:
self.classifiers[category] = deepcopy(self.classifier_prototype) self.classifiers[category] = deepcopy(self.classifier_prototype)
self.classifiers[category].fit(x[category], y[category]) self.classifiers[category].fit(x[category], y[category])
@@ -121,8 +128,10 @@ class StaticLazyConstraintsComponent(Component):
x[category] = [] x[category] = []
constraints[category] = [] constraints[category] = []
x[category] += [instance.get_constraint_features(cid)] x[category] += [instance.get_constraint_features(cid)]
c = LazyConstraint(cid=cid, c = LazyConstraint(
obj=solver.internal_solver.extract_constraint(cid)) cid=cid,
obj=solver.internal_solver.extract_constraint(cid),
)
constraints[category] += [c] constraints[category] += [c]
self.pool.append(c) self.pool.append(c)
logger.info("%8d lazy constraints extracted" % len(self.pool)) logger.info("%8d lazy constraints extracted" % len(self.pool))
@@ -141,7 +150,13 @@ class StaticLazyConstraintsComponent(Component):
self.pool.remove(c) self.pool.remove(c)
solver.internal_solver.add_constraint(c.obj) solver.internal_solver.add_constraint(c.obj)
instance.found_violated_lazy_constraints += [c.cid] instance.found_violated_lazy_constraints += [c.cid]
logger.info("%8d lazy constraints added %8d in the pool" % (n_added, len(self.pool))) logger.info(
"%8d lazy constraints added %8d in the pool"
% (
n_added,
len(self.pool),
)
)
def _collect_constraints(self, train_instances): def _collect_constraints(self, train_instances):
constraints = {} constraints = {}

21
miplearn/components/objective.py
View File

@@ -1,13 +1,20 @@
# MIPLearn: Extensible Framework for Learning-Enhanced Mixed-Integer Optimization # MIPLearn: Extensible Framework for Learning-Enhanced Mixed-Integer Optimization
# Copyright (C) 2020, UChicago Argonne, LLC. All rights reserved. # Copyright (C) 2020, UChicago Argonne, LLC. All rights reserved.
# Released under the modified BSD license. See COPYING.md for more details. # Released under the modified BSD license. See COPYING.md for more details.
from sklearn.metrics import mean_squared_error, explained_variance_score, max_error, mean_absolute_error, r2_score from sklearn.metrics import (
mean_squared_error,
explained_variance_score,
max_error,
mean_absolute_error,
r2_score,
)
from .. import Component, InstanceFeaturesExtractor, ObjectiveValueExtractor from .. import Component, InstanceFeaturesExtractor, ObjectiveValueExtractor
from sklearn.linear_model import LinearRegression from sklearn.linear_model import LinearRegression
from copy import deepcopy from copy import deepcopy
import numpy as np import numpy as np
import logging import logging
logger = logging.getLogger(__name__) logger = logging.getLogger(__name__)
@@ -15,12 +22,12 @@ class ObjectiveValueComponent(Component):
""" """
A Component which predicts the optimal objective value of the problem. A Component which predicts the optimal objective value of the problem.
""" """
def __init__(self,
regressor=LinearRegression()): def __init__(self, regressor=LinearRegression()):
self.ub_regressor = None self.ub_regressor = None
self.lb_regressor = None self.lb_regressor = None
self.regressor_prototype = regressor self.regressor_prototype = regressor
def before_solve(self, solver, instance, model): def before_solve(self, solver, instance, model):
if self.ub_regressor is not None: if self.ub_regressor is not None:
logger.info("Predicting optimal value...") logger.info("Predicting optimal value...")
@@ -28,7 +35,7 @@ class ObjectiveValueComponent(Component):
instance.predicted_ub = ub instance.predicted_ub = ub
instance.predicted_lb = lb instance.predicted_lb = lb
logger.info("Predicted values: lb=%.2f, ub=%.2f" % (lb, ub)) logger.info("Predicted values: lb=%.2f, ub=%.2f" % (lb, ub))
def after_solve(self, solver, instance, model, results): def after_solve(self, solver, instance, model, results):
if self.ub_regressor is not None: if self.ub_regressor is not None:
results["Predicted UB"] = instance.predicted_ub results["Predicted UB"] = instance.predicted_ub
@@ -36,7 +43,7 @@ class ObjectiveValueComponent(Component):
else: else:
results["Predicted UB"] = None results["Predicted UB"] = None
results["Predicted LB"] = None results["Predicted LB"] = None
def fit(self, training_instances): def fit(self, training_instances):
logger.debug("Extracting features...") logger.debug("Extracting features...")
features = InstanceFeaturesExtractor().extract(training_instances) features = InstanceFeaturesExtractor().extract(training_instances)
@@ -50,7 +57,7 @@ class ObjectiveValueComponent(Component):
self.ub_regressor.fit(features, ub.ravel()) self.ub_regressor.fit(features, ub.ravel())
logger.debug("Fitting ub_regressor...") logger.debug("Fitting ub_regressor...")
self.lb_regressor.fit(features, lb.ravel()) self.lb_regressor.fit(features, lb.ravel())
def predict(self, instances): def predict(self, instances):
features = InstanceFeaturesExtractor().extract(instances) features = InstanceFeaturesExtractor().extract(instances)
lb = self.lb_regressor.predict(features) lb = self.lb_regressor.predict(features)

50
miplearn/components/primal.py
View File

@@ -19,10 +19,12 @@ class PrimalSolutionComponent(Component):
A component that predicts primal solutions. A component that predicts primal solutions.
""" """
def __init__(self, def __init__(
classifier=AdaptiveClassifier(), self,
mode="exact", classifier=AdaptiveClassifier(),
threshold=MinPrecisionThreshold(0.98)): mode="exact",
threshold=MinPrecisionThreshold(0.98),
):
self.mode = mode self.mode = mode
self.classifiers = {} self.classifiers = {}
self.thresholds = {} self.thresholds = {}
@@ -51,9 +53,10 @@ class PrimalSolutionComponent(Component):
features = VariableFeaturesExtractor().extract(training_instances) features = VariableFeaturesExtractor().extract(training_instances)
solutions = SolutionExtractor().extract(training_instances) solutions = SolutionExtractor().extract(training_instances)
for category in tqdm(features.keys(), for category in tqdm(
desc="Fit (primal)", features.keys(),
): desc="Fit (primal)",
):
x_train = features[category] x_train = features[category]
for label in [0, 1]: for label in [0, 1]:
y_train = solutions[category][:, label].astype(int) y_train = solutions[category][:, label].astype(int)
@@ -74,9 +77,15 @@ class PrimalSolutionComponent(Component):
# Find threshold (dynamic or static) # Find threshold (dynamic or static)
if isinstance(self.threshold_prototype, DynamicThreshold): if isinstance(self.threshold_prototype, DynamicThreshold):
self.thresholds[category, label] = self.threshold_prototype.find(clf, x_train, y_train) self.thresholds[category, label] = self.threshold_prototype.find(
clf,
x_train,
y_train,
)
else: else:
self.thresholds[category, label] = deepcopy(self.threshold_prototype) self.thresholds[category, label] = deepcopy(
self.threshold_prototype
)
self.classifiers[category, label] = clf self.classifiers[category, label] = clf
@@ -98,18 +107,21 @@ class PrimalSolutionComponent(Component):
ws = np.array([[1 - clf, clf] for _ in range(n)]) ws = np.array([[1 - clf, clf] for _ in range(n)])
else: else:
ws = clf.predict_proba(x_test[category]) ws = clf.predict_proba(x_test[category])
assert ws.shape == (n, 2), "ws.shape should be (%d, 2) not %s" % (n, ws.shape) assert ws.shape == (n, 2), "ws.shape should be (%d, 2) not %s" % (
n,
ws.shape,
)
for (i, (var, index)) in enumerate(var_split[category]): for (i, (var, index)) in enumerate(var_split[category]):
if ws[i, 1] >= self.thresholds[category, label]: if ws[i, 1] >= self.thresholds[category, label]:
solution[var][index] = label solution[var][index] = label
return solution return solution
def evaluate(self, instances): def evaluate(self, instances):
ev = {"Fix zero": {}, ev = {"Fix zero": {}, "Fix one": {}}
"Fix one": {}} for instance_idx in tqdm(
for instance_idx in tqdm(range(len(instances)), range(len(instances)),
desc="Evaluate (primal)", desc="Evaluate (primal)",
): ):
instance = instances[instance_idx] instance = instances[instance_idx]
solution_actual = instance.solution solution_actual = instance.solution
solution_pred = self.predict(instance) solution_pred = self.predict(instance)
@@ -143,6 +155,10 @@ class PrimalSolutionComponent(Component):
tn_one = len(pred_one_negative & vars_zero) tn_one = len(pred_one_negative & vars_zero)
fn_one = len(pred_one_negative & vars_one) fn_one = len(pred_one_negative & vars_one)
ev["Fix zero"][instance_idx] = classifier_evaluation_dict(tp_zero, tn_zero, fp_zero, fn_zero) ev["Fix zero"][instance_idx] = classifier_evaluation_dict(
ev["Fix one"][instance_idx] = classifier_evaluation_dict(tp_one, tn_one, fp_one, fn_one) tp_zero, tn_zero, fp_zero, fn_zero
)
ev["Fix one"][instance_idx] = classifier_evaluation_dict(
tp_one, tn_one, fp_one, fn_one
)
return ev return ev

67
miplearn/components/relaxation.py
View File

@@ -51,14 +51,15 @@ class RelaxationComponent(Component):
If `check_dropped` is true, set the maximum number of iterations in the lazy constraint loop. If `check_dropped` is true, set the maximum number of iterations in the lazy constraint loop.
""" """
def __init__(self, def __init__(
classifier=CountingClassifier(), self,
threshold=0.95, classifier=CountingClassifier(),
slack_tolerance=1e-5, threshold=0.95,
check_dropped=False, slack_tolerance=1e-5,
violation_tolerance=1e-5, check_dropped=False,
max_iterations=3, violation_tolerance=1e-5,
): max_iterations=3,
):
self.classifiers = {} self.classifiers = {}
self.classifier_prototype = classifier self.classifier_prototype = classifier
self.threshold = threshold self.threshold = threshold
@@ -77,16 +78,20 @@ class RelaxationComponent(Component):
logger.info("Predicting redundant LP constraints...") logger.info("Predicting redundant LP constraints...")
cids = solver.internal_solver.get_constraint_ids() cids = solver.internal_solver.get_constraint_ids()
x, constraints = self.x([instance], x, constraints = self.x(
constraint_ids=cids, [instance],
return_constraints=True) constraint_ids=cids,
return_constraints=True,
)
y = self.predict(x) y = self.predict(x)
for category in y.keys(): for category in y.keys():
for i in range(len(y[category])): for i in range(len(y[category])):
if y[category][i][0] == 1: if y[category][i][0] == 1:
cid = constraints[category][i] cid = constraints[category][i]
c = LazyConstraint(cid=cid, c = LazyConstraint(
obj=solver.internal_solver.extract_constraint(cid)) cid=cid,
obj=solver.internal_solver.extract_constraint(cid),
)
self.pool += [c] self.pool += [c]
logger.info("Extracted %d predicted constraints" % len(self.pool)) logger.info("Extracted %d predicted constraints" % len(self.pool))
@@ -98,21 +103,19 @@ class RelaxationComponent(Component):
x = self.x(training_instances) x = self.x(training_instances)
y = self.y(training_instances) y = self.y(training_instances)
logger.debug("Fitting...") logger.debug("Fitting...")
for category in tqdm(x.keys(), for category in tqdm(x.keys(), desc="Fit (relaxation)"):
desc="Fit (relaxation)"):
if category not in self.classifiers: if category not in self.classifiers:
self.classifiers[category] = deepcopy(self.classifier_prototype) self.classifiers[category] = deepcopy(self.classifier_prototype)
self.classifiers[category].fit(x[category], y[category]) self.classifiers[category].fit(x[category], y[category])
def x(self, def x(self, instances, constraint_ids=None, return_constraints=False):
instances,
constraint_ids=None,
return_constraints=False):
x = {} x = {}
constraints = {} constraints = {}
for instance in tqdm(InstanceIterator(instances), for instance in tqdm(
desc="Extract (relaxation:x)", InstanceIterator(instances),
disable=len(instances) < 5): desc="Extract (relaxation:x)",
disable=len(instances) < 5,
):
if constraint_ids is not None: if constraint_ids is not None:
cids = constraint_ids cids = constraint_ids
else: else:
@@ -133,9 +136,11 @@ class RelaxationComponent(Component):
def y(self, instances): def y(self, instances):
y = {} y = {}
for instance in tqdm(InstanceIterator(instances), for instance in tqdm(
desc="Extract (relaxation:y)", InstanceIterator(instances),
disable=len(instances) < 5): desc="Extract (relaxation:y)",
disable=len(instances) < 5,
):
for (cid, slack) in instance.slacks.items(): for (cid, slack) in instance.slacks.items():
category = instance.get_constraint_category(cid) category = instance.get_constraint_category(cid)
if category is None: if category is None:
@@ -154,7 +159,7 @@ class RelaxationComponent(Component):
if category not in self.classifiers: if category not in self.classifiers:
continue continue
y[category] = [] y[category] = []
#x_cat = np.array(x_cat) # x_cat = np.array(x_cat)
proba = self.classifiers[category].predict_proba(x_cat) proba = self.classifiers[category].predict_proba(x_cat)
for i in range(len(proba)): for i in range(len(proba)):
if proba[i][1] >= self.threshold: if proba[i][1] >= self.threshold:
@@ -191,13 +196,19 @@ class RelaxationComponent(Component):
logger.debug("Checking that dropped constraints are satisfied...") logger.debug("Checking that dropped constraints are satisfied...")
constraints_to_add = [] constraints_to_add = []
for c in self.pool: for c in self.pool:
if not solver.internal_solver.is_constraint_satisfied(c.obj, self.violation_tolerance): if not solver.internal_solver.is_constraint_satisfied(
c.obj,
self.violation_tolerance,
):
constraints_to_add.append(c) constraints_to_add.append(c)
for c in constraints_to_add: for c in constraints_to_add:
self.pool.remove(c) self.pool.remove(c)
solver.internal_solver.add_constraint(c.obj) solver.internal_solver.add_constraint(c.obj)
if len(constraints_to_add) > 0: if len(constraints_to_add) > 0:
logger.info("%8d constraints %8d in the pool" % (len(constraints_to_add), len(self.pool))) logger.info(
"%8d constraints %8d in the pool"
% (len(constraints_to_add), len(self.pool))
)
return True return True
else: else:
return False return False

28
miplearn/components/tests/test_lazy_dynamic.py
View File

@@ -28,9 +28,9 @@ def test_lazy_fit():
assert "c" in component.classifiers assert "c" in component.classifiers
# Should provide correct x_train to each classifier # Should provide correct x_train to each classifier
expected_x_train_a = np.array([[67., 21.75, 1287.92], [70., 23.75, 1199.83]]) expected_x_train_a = np.array([[67.0, 21.75, 1287.92], [70.0, 23.75, 1199.83]])
expected_x_train_b = np.array([[67., 21.75, 1287.92], [70., 23.75, 1199.83]]) expected_x_train_b = np.array([[67.0, 21.75, 1287.92], [70.0, 23.75, 1199.83]])
expected_x_train_c = np.array([[67., 21.75, 1287.92], [70., 23.75, 1199.83]]) expected_x_train_c = np.array([[67.0, 21.75, 1287.92], [70.0, 23.75, 1199.83]])
actual_x_train_a = component.classifiers["a"].fit.call_args[0][0] actual_x_train_a = component.classifiers["a"].fit.call_args[0][0]
actual_x_train_b = component.classifiers["b"].fit.call_args[0][0] actual_x_train_b = component.classifiers["b"].fit.call_args[0][0]
actual_x_train_c = component.classifiers["c"].fit.call_args[0][0] actual_x_train_c = component.classifiers["c"].fit.call_args[0][0]
@@ -56,16 +56,15 @@ def test_lazy_before():
solver = LearningSolver() solver = LearningSolver()
solver.internal_solver = Mock(spec=InternalSolver) solver.internal_solver = Mock(spec=InternalSolver)
component = DynamicLazyConstraintsComponent(threshold=0.10) component = DynamicLazyConstraintsComponent(threshold=0.10)
component.classifiers = {"a": Mock(spec=Classifier), component.classifiers = {"a": Mock(spec=Classifier), "b": Mock(spec=Classifier)}
"b": Mock(spec=Classifier)}
component.classifiers["a"].predict_proba = Mock(return_value=[[0.95, 0.05]]) component.classifiers["a"].predict_proba = Mock(return_value=[[0.95, 0.05]])
component.classifiers["b"].predict_proba = Mock(return_value=[[0.02, 0.80]]) component.classifiers["b"].predict_proba = Mock(return_value=[[0.02, 0.80]])
component.before_solve(solver, instances[0], models[0]) component.before_solve(solver, instances[0], models[0])
# Should ask classifier likelihood of each constraint being violated # Should ask classifier likelihood of each constraint being violated
expected_x_test_a = np.array([[67., 21.75, 1287.92]]) expected_x_test_a = np.array([[67.0, 21.75, 1287.92]])
expected_x_test_b = np.array([[67., 21.75, 1287.92]]) expected_x_test_b = np.array([[67.0, 21.75, 1287.92]])
actual_x_test_a = component.classifiers["a"].predict_proba.call_args[0][0] actual_x_test_a = component.classifiers["a"].predict_proba.call_args[0][0]
actual_x_test_b = component.classifiers["b"].predict_proba.call_args[0][0] actual_x_test_b = component.classifiers["b"].predict_proba.call_args[0][0]
assert norm(expected_x_test_a - actual_x_test_a) < E assert norm(expected_x_test_a - actual_x_test_a) < E
@@ -82,13 +81,15 @@ def test_lazy_before():
def test_lazy_evaluate(): def test_lazy_evaluate():
instances, models = get_test_pyomo_instances() instances, models = get_test_pyomo_instances()
component = DynamicLazyConstraintsComponent() component = DynamicLazyConstraintsComponent()
component.classifiers = {"a": Mock(spec=Classifier), component.classifiers = {
"b": Mock(spec=Classifier), "a": Mock(spec=Classifier),
"c": Mock(spec=Classifier)} "b": Mock(spec=Classifier),
"c": Mock(spec=Classifier),
}
component.classifiers["a"].predict_proba = Mock(return_value=[[1.0, 0.0]]) component.classifiers["a"].predict_proba = Mock(return_value=[[1.0, 0.0]])
component.classifiers["b"].predict_proba = Mock(return_value=[[0.0, 1.0]]) component.classifiers["b"].predict_proba = Mock(return_value=[[0.0, 1.0]])
component.classifiers["c"].predict_proba = Mock(return_value=[[0.0, 1.0]]) component.classifiers["c"].predict_proba = Mock(return_value=[[0.0, 1.0]])
instances[0].found_violated_lazy_constraints = ["a", "b", "c"] instances[0].found_violated_lazy_constraints = ["a", "b", "c"]
instances[1].found_violated_lazy_constraints = ["b", "d"] instances[1].found_violated_lazy_constraints = ["b", "d"]
assert component.evaluate(instances) == { assert component.evaluate(instances) == {
@@ -96,7 +97,7 @@ def test_lazy_evaluate():
"Accuracy": 0.75, "Accuracy": 0.75,
"F1 score": 0.8, "F1 score": 0.8,
"Precision": 1.0, "Precision": 1.0,
"Recall": 2/3., "Recall": 2 / 3.0,
"Predicted positive": 2, "Predicted positive": 2,
"Predicted negative": 2, "Predicted negative": 2,
"Condition positive": 3, "Condition positive": 3,
@@ -135,6 +136,5 @@ def test_lazy_evaluate():
"False positive (%)": 25.0, "False positive (%)": 25.0,
"True negative (%)": 25.0, "True negative (%)": 25.0,
"True positive (%)": 25.0, "True positive (%)": 25.0,
} },
} }

202
miplearn/components/tests/test_lazy_static.py
View File

@@ -4,10 +4,12 @@
from unittest.mock import Mock, call from unittest.mock import Mock, call
from miplearn import (StaticLazyConstraintsComponent, from miplearn import (
LearningSolver, StaticLazyConstraintsComponent,
Instance, LearningSolver,
InternalSolver) Instance,
InternalSolver,
)
from miplearn.classifiers import Classifier from miplearn.classifiers import Classifier
@@ -23,39 +25,47 @@ def test_usage_with_solver():
instance = Mock(spec=Instance) instance = Mock(spec=Instance)
instance.has_static_lazy_constraints = Mock(return_value=True) instance.has_static_lazy_constraints = Mock(return_value=True)
instance.is_constraint_lazy = Mock(side_effect=lambda cid: { instance.is_constraint_lazy = Mock(
"c1": False, side_effect=lambda cid: {
"c2": True, "c1": False,
"c3": True, "c2": True,
"c4": True, "c3": True,
}[cid]) "c4": True,
instance.get_constraint_features = Mock(side_effect=lambda cid: { }[cid]
"c2": [1.0, 0.0], )
"c3": [0.5, 0.5], instance.get_constraint_features = Mock(
"c4": [1.0], side_effect=lambda cid: {
}[cid]) "c2": [1.0, 0.0],
instance.get_constraint_category = Mock(side_effect=lambda cid: { "c3": [0.5, 0.5],
"c2": "type-a", "c4": [1.0],
"c3": "type-a", }[cid]
"c4": "type-b", )
}[cid]) instance.get_constraint_category = Mock(
side_effect=lambda cid: {
"c2": "type-a",
"c3": "type-a",
"c4": "type-b",
}[cid]
)
component = StaticLazyConstraintsComponent(threshold=0.90, component = StaticLazyConstraintsComponent(
use_two_phase_gap=False, threshold=0.90, use_two_phase_gap=False, violation_tolerance=1.0
violation_tolerance=1.0) )
component.classifiers = { component.classifiers = {
"type-a": Mock(spec=Classifier), "type-a": Mock(spec=Classifier),
"type-b": Mock(spec=Classifier), "type-b": Mock(spec=Classifier),
} }
component.classifiers["type-a"].predict_proba = \ component.classifiers["type-a"].predict_proba = Mock(
Mock(return_value=[ return_value=[
[0.20, 0.80], [0.20, 0.80],
[0.05, 0.95], [0.05, 0.95],
]) ]
component.classifiers["type-b"].predict_proba = \ )
Mock(return_value=[ component.classifiers["type-b"].predict_proba = Mock(
return_value=[
[0.02, 0.98], [0.02, 0.98],
]) ]
)
# LearningSolver calls before_solve # LearningSolver calls before_solve
component.before_solve(solver, instance, None) component.before_solve(solver, instance, None)
@@ -67,37 +77,59 @@ def test_usage_with_solver():
internal.get_constraint_ids.assert_called_once() internal.get_constraint_ids.assert_called_once()
# Should ask if each constraint in the model is lazy # Should ask if each constraint in the model is lazy
instance.is_constraint_lazy.assert_has_calls([ instance.is_constraint_lazy.assert_has_calls(
call("c1"), call("c2"), call("c3"), call("c4"), [
]) call("c1"),
call("c2"),
call("c3"),
call("c4"),
]
)
# For the lazy ones, should ask for features # For the lazy ones, should ask for features
instance.get_constraint_features.assert_has_calls([ instance.get_constraint_features.assert_has_calls(
call("c2"), call("c3"), call("c4"), [
]) call("c2"),
call("c3"),
call("c4"),
]
)
# Should also ask for categories # Should also ask for categories
assert instance.get_constraint_category.call_count == 3 assert instance.get_constraint_category.call_count == 3
instance.get_constraint_category.assert_has_calls([ instance.get_constraint_category.assert_has_calls(
call("c2"), call("c3"), call("c4"), [
]) call("c2"),
call("c3"),
call("c4"),
]
)
# Should ask internal solver to remove constraints identified as lazy # Should ask internal solver to remove constraints identified as lazy
assert internal.extract_constraint.call_count == 3 assert internal.extract_constraint.call_count == 3
internal.extract_constraint.assert_has_calls([ internal.extract_constraint.assert_has_calls(
call("c2"), call("c3"), call("c4"), [
]) call("c2"),
call("c3"),
call("c4"),
]
)
# Should ask ML to predict whether each lazy constraint should be enforced # Should ask ML to predict whether each lazy constraint should be enforced
component.classifiers["type-a"].predict_proba.assert_called_once_with([[1.0, 0.0], [0.5, 0.5]]) component.classifiers["type-a"].predict_proba.assert_called_once_with(
[[1.0, 0.0], [0.5, 0.5]]
)
component.classifiers["type-b"].predict_proba.assert_called_once_with([[1.0]]) component.classifiers["type-b"].predict_proba.assert_called_once_with([[1.0]])
# For the ones that should be enforced, should ask solver to re-add them # For the ones that should be enforced, should ask solver to re-add them
# to the formulation. The remaining ones should remain in the pool. # to the formulation. The remaining ones should remain in the pool.
assert internal.add_constraint.call_count == 2 assert internal.add_constraint.call_count == 2
internal.add_constraint.assert_has_calls([ internal.add_constraint.assert_has_calls(
call("<c3>"), call("<c4>"), [
]) call("<c3>"),
call("<c4>"),
]
)
internal.add_constraint.reset_mock() internal.add_constraint.reset_mock()
# LearningSolver calls after_iteration (first time) # LearningSolver calls after_iteration (first time)
@@ -126,37 +158,45 @@ def test_usage_with_solver():
def test_fit(): def test_fit():
instance_1 = Mock(spec=Instance) instance_1 = Mock(spec=Instance)
instance_1.found_violated_lazy_constraints = ["c1", "c2", "c4", "c5"] instance_1.found_violated_lazy_constraints = ["c1", "c2", "c4", "c5"]
instance_1.get_constraint_category = Mock(side_effect=lambda cid: { instance_1.get_constraint_category = Mock(
"c1": "type-a", side_effect=lambda cid: {
"c2": "type-a", "c1": "type-a",
"c3": "type-a", "c2": "type-a",
"c4": "type-b", "c3": "type-a",
"c5": "type-b", "c4": "type-b",
}[cid]) "c5": "type-b",
instance_1.get_constraint_features = Mock(side_effect=lambda cid: { }[cid]
"c1": [1, 1], )
"c2": [1, 2], instance_1.get_constraint_features = Mock(
"c3": [1, 3], side_effect=lambda cid: {
"c4": [1, 4, 0], "c1": [1, 1],
"c5": [1, 5, 0], "c2": [1, 2],
}[cid]) "c3": [1, 3],
"c4": [1, 4, 0],
"c5": [1, 5, 0],
}[cid]
)
instance_2 = Mock(spec=Instance) instance_2 = Mock(spec=Instance)
instance_2.found_violated_lazy_constraints = ["c2", "c3", "c4"] instance_2.found_violated_lazy_constraints = ["c2", "c3", "c4"]
instance_2.get_constraint_category = Mock(side_effect=lambda cid: { instance_2.get_constraint_category = Mock(
"c1": "type-a", side_effect=lambda cid: {
"c2": "type-a", "c1": "type-a",
"c3": "type-a", "c2": "type-a",
"c4": "type-b", "c3": "type-a",
"c5": "type-b", "c4": "type-b",
}[cid]) "c5": "type-b",
instance_2.get_constraint_features = Mock(side_effect=lambda cid: { }[cid]
"c1": [2, 1], )
"c2": [2, 2], instance_2.get_constraint_features = Mock(
"c3": [2, 3], side_effect=lambda cid: {
"c4": [2, 4, 0], "c1": [2, 1],
"c5": [2, 5, 0], "c2": [2, 2],
}[cid]) "c3": [2, 3],
"c4": [2, 4, 0],
"c5": [2, 5, 0],
}[cid]
)
instances = [instance_1, instance_2] instances = [instance_1, instance_2]
component = StaticLazyConstraintsComponent() component = StaticLazyConstraintsComponent()
@@ -171,18 +211,22 @@ def test_fit():
} }
expected_x = { expected_x = {
"type-a": [[1, 1], [1, 2], [1, 3], [2, 1], [2, 2], [2, 3]], "type-a": [[1, 1], [1, 2], [1, 3], [2, 1], [2, 2], [2, 3]],
"type-b": [[1, 4, 0], [1, 5, 0], [2, 4, 0], [2, 5, 0]] "type-b": [[1, 4, 0], [1, 5, 0], [2, 4, 0], [2, 5, 0]],
} }
expected_y = { expected_y = {
"type-a": [[0, 1], [0, 1], [1, 0], [1, 0], [0, 1], [0, 1]], "type-a": [[0, 1], [0, 1], [1, 0], [1, 0], [0, 1], [0, 1]],
"type-b": [[0, 1], [0, 1], [0, 1], [1, 0]] "type-b": [[0, 1], [0, 1], [0, 1], [1, 0]],
} }
assert component._collect_constraints(instances) == expected_constraints assert component._collect_constraints(instances) == expected_constraints
assert component.x(instances) == expected_x assert component.x(instances) == expected_x
assert component.y(instances) == expected_y assert component.y(instances) == expected_y
component.fit(instances) component.fit(instances)
component.classifiers["type-a"].fit.assert_called_once_with(expected_x["type-a"], component.classifiers["type-a"].fit.assert_called_once_with(
expected_y["type-a"]) expected_x["type-a"],
component.classifiers["type-b"].fit.assert_called_once_with(expected_x["type-b"], expected_y["type-a"],
expected_y["type-b"]) )
component.classifiers["type-b"].fit.assert_called_once_with(
expected_x["type-b"],
expected_y["type-b"],
)

36
miplearn/components/tests/test_objective.py
View File

@@ -16,8 +16,10 @@ def test_usage():
comp.fit(instances) comp.fit(instances)
assert instances[0].lower_bound == 1183.0 assert instances[0].lower_bound == 1183.0
assert instances[0].upper_bound == 1183.0 assert instances[0].upper_bound == 1183.0
assert np.round(comp.predict(instances), 2).tolist() == [[1183.0, 1183.0], assert np.round(comp.predict(instances), 2).tolist() == [
[1070.0, 1070.0]] [1183.0, 1183.0],
[1070.0, 1070.0],
]
def test_obj_evaluate(): def test_obj_evaluate():
@@ -28,20 +30,20 @@ def test_obj_evaluate():
comp.fit(instances) comp.fit(instances)
ev = comp.evaluate(instances) ev = comp.evaluate(instances)
assert ev == { assert ev == {
'Lower bound': { "Lower bound": {
'Explained variance': 0.0, "Explained variance": 0.0,
'Max error': 183.0, "Max error": 183.0,
'Mean absolute error': 126.5, "Mean absolute error": 126.5,
'Mean squared error': 19194.5, "Mean squared error": 19194.5,
'Median absolute error': 126.5, "Median absolute error": 126.5,
'R2': -5.012843605607331, "R2": -5.012843605607331,
},
"Upper bound": {
"Explained variance": 0.0,
"Max error": 183.0,
"Mean absolute error": 126.5,
"Mean squared error": 19194.5,
"Median absolute error": 126.5,
"R2": -5.012843605607331,
}, },
'Upper bound': {
'Explained variance': 0.0,
'Max error': 183.0,
'Mean absolute error': 126.5,
'Mean squared error': 19194.5,
'Median absolute error': 126.5,
'R2': -5.012843605607331,
}
} }

135
miplearn/components/tests/test_primal.py
View File

@@ -25,71 +25,82 @@ def test_predict():
def test_evaluate(): def test_evaluate():
instances, models = get_test_pyomo_instances() instances, models = get_test_pyomo_instances()
clf_zero = Mock(spec=Classifier) clf_zero = Mock(spec=Classifier)
clf_zero.predict_proba = Mock(return_value=np.array([ clf_zero.predict_proba = Mock(
[0., 1.], # x[0] return_value=np.array(
[0., 1.], # x[1] [
[1., 0.], # x[2] [0.0, 1.0], # x[0]
[1., 0.], # x[3] [0.0, 1.0], # x[1]
])) [1.0, 0.0], # x[2]
[1.0, 0.0], # x[3]
]
)
)
clf_one = Mock(spec=Classifier) clf_one = Mock(spec=Classifier)
clf_one.predict_proba = Mock(return_value=np.array([ clf_one.predict_proba = Mock(
[1., 0.], # x[0] instances[0] return_value=np.array(
[1., 0.], # x[1] instances[0] [
[0., 1.], # x[2] instances[0] [1.0, 0.0], # x[0] instances[0]
[1., 0.], # x[3] instances[0] [1.0, 0.0], # x[1] instances[0]
])) [0.0, 1.0], # x[2] instances[0]
comp = PrimalSolutionComponent(classifier=[clf_zero, clf_one], [1.0, 0.0], # x[3] instances[0]
threshold=0.50) ]
)
)
comp = PrimalSolutionComponent(classifier=[clf_zero, clf_one], threshold=0.50)
comp.fit(instances[:1]) comp.fit(instances[:1])
assert comp.predict(instances[0]) == {"x": {0: 0, assert comp.predict(instances[0]) == {"x": {0: 0, 1: 0, 2: 1, 3: None}}
1: 0, assert instances[0].solution == {"x": {0: 1, 1: 0, 2: 1, 3: 1}}
2: 1,
3: None}}
assert instances[0].solution == {"x": {0: 1,
1: 0,
2: 1,
3: 1}}
ev = comp.evaluate(instances[:1]) ev = comp.evaluate(instances[:1])
assert ev == {'Fix one': {0: {'Accuracy': 0.5, assert ev == {
'Condition negative': 1, "Fix one": {
'Condition negative (%)': 25.0, 0: {
'Condition positive': 3, "Accuracy": 0.5,
'Condition positive (%)': 75.0, "Condition negative": 1,
'F1 score': 0.5, "Condition negative (%)": 25.0,
'False negative': 2, "Condition positive": 3,
'False negative (%)': 50.0, "Condition positive (%)": 75.0,
'False positive': 0, "F1 score": 0.5,
'False positive (%)': 0.0, "False negative": 2,
'Precision': 1.0, "False negative (%)": 50.0,
'Predicted negative': 3, "False positive": 0,
'Predicted negative (%)': 75.0, "False positive (%)": 0.0,
'Predicted positive': 1, "Precision": 1.0,
'Predicted positive (%)': 25.0, "Predicted negative": 3,
'Recall': 0.3333333333333333, "Predicted negative (%)": 75.0,
'True negative': 1, "Predicted positive": 1,
'True negative (%)': 25.0, "Predicted positive (%)": 25.0,
'True positive': 1, "Recall": 0.3333333333333333,
'True positive (%)': 25.0}}, "True negative": 1,
'Fix zero': {0: {'Accuracy': 0.75, "True negative (%)": 25.0,
'Condition negative': 3, "True positive": 1,
'Condition negative (%)': 75.0, "True positive (%)": 25.0,
'Condition positive': 1, }
'Condition positive (%)': 25.0, },
'F1 score': 0.6666666666666666, "Fix zero": {
'False negative': 0, 0: {
'False negative (%)': 0.0, "Accuracy": 0.75,
'False positive': 1, "Condition negative": 3,
'False positive (%)': 25.0, "Condition negative (%)": 75.0,
'Precision': 0.5, "Condition positive": 1,
'Predicted negative': 2, "Condition positive (%)": 25.0,
'Predicted negative (%)': 50.0, "F1 score": 0.6666666666666666,
'Predicted positive': 2, "False negative": 0,
'Predicted positive (%)': 50.0, "False negative (%)": 0.0,
'Recall': 1.0, "False positive": 1,
'True negative': 2, "False positive (%)": 25.0,
'True negative (%)': 50.0, "Precision": 0.5,
'True positive': 1, "Predicted negative": 2,
'True positive (%)': 25.0}}} "Predicted negative (%)": 50.0,
"Predicted positive": 2,
"Predicted positive (%)": 50.0,
"Recall": 1.0,
"True negative": 2,
"True negative (%)": 50.0,
"True positive": 1,
"True positive (%)": 25.0,
}
},
}
def test_primal_parallel_fit(): def test_primal_parallel_fit():

208
miplearn/components/tests/test_relaxation.py
View File

@@ -4,10 +4,7 @@
from unittest.mock import Mock, call from unittest.mock import Mock, call
from miplearn import (RelaxationComponent, from miplearn import RelaxationComponent, LearningSolver, Instance, InternalSolver
LearningSolver,
Instance,
InternalSolver)
from miplearn.classifiers import Classifier from miplearn.classifiers import Classifier
@@ -16,41 +13,49 @@ def _setup():
internal = solver.internal_solver = Mock(spec=InternalSolver) internal = solver.internal_solver = Mock(spec=InternalSolver)
internal.get_constraint_ids = Mock(return_value=["c1", "c2", "c3", "c4"]) internal.get_constraint_ids = Mock(return_value=["c1", "c2", "c3", "c4"])
internal.get_constraint_slacks = Mock(side_effect=lambda: { internal.get_constraint_slacks = Mock(
"c1": 0.5, side_effect=lambda: {
"c2": 0.0, "c1": 0.5,
"c3": 0.0, "c2": 0.0,
"c4": 1.4, "c3": 0.0,
}) "c4": 1.4,
}
)
internal.extract_constraint = Mock(side_effect=lambda cid: "<%s>" % cid) internal.extract_constraint = Mock(side_effect=lambda cid: "<%s>" % cid)
internal.is_constraint_satisfied = Mock(return_value=False) internal.is_constraint_satisfied = Mock(return_value=False)
instance = Mock(spec=Instance) instance = Mock(spec=Instance)
instance.get_constraint_features = Mock(side_effect=lambda cid: { instance.get_constraint_features = Mock(
"c2": [1.0, 0.0], side_effect=lambda cid: {
"c3": [0.5, 0.5], "c2": [1.0, 0.0],
"c4": [1.0], "c3": [0.5, 0.5],
}[cid]) "c4": [1.0],
instance.get_constraint_category = Mock(side_effect=lambda cid: { }[cid]
"c1": None, )
"c2": "type-a", instance.get_constraint_category = Mock(
"c3": "type-a", side_effect=lambda cid: {
"c4": "type-b", "c1": None,
}[cid]) "c2": "type-a",
"c3": "type-a",
"c4": "type-b",
}[cid]
)
classifiers = { classifiers = {
"type-a": Mock(spec=Classifier), "type-a": Mock(spec=Classifier),
"type-b": Mock(spec=Classifier), "type-b": Mock(spec=Classifier),
} }
classifiers["type-a"].predict_proba = \ classifiers["type-a"].predict_proba = Mock(
Mock(return_value=[ return_value=[
[0.20, 0.80], [0.20, 0.80],
[0.05, 0.95], [0.05, 0.95],
]) ]
classifiers["type-b"].predict_proba = \ )
Mock(return_value=[ classifiers["type-b"].predict_proba = Mock(
return_value=[
[0.02, 0.98], [0.02, 0.98],
]) ]
)
return solver, internal, instance, classifiers return solver, internal, instance, classifiers
@@ -72,25 +77,39 @@ def test_usage():
# Should query category and features for each constraint in the model # Should query category and features for each constraint in the model
assert instance.get_constraint_category.call_count == 4 assert instance.get_constraint_category.call_count == 4
instance.get_constraint_category.assert_has_calls([ instance.get_constraint_category.assert_has_calls(
call("c1"), call("c2"), call("c3"), call("c4"), [
]) call("c1"),
call("c2"),
call("c3"),
call("c4"),
]
)
# For constraint with non-null categories, should ask for features # For constraint with non-null categories, should ask for features
assert instance.get_constraint_features.call_count == 3 assert instance.get_constraint_features.call_count == 3
instance.get_constraint_features.assert_has_calls([ instance.get_constraint_features.assert_has_calls(
call("c2"), call("c3"), call("c4"), [
]) call("c2"),
call("c3"),
call("c4"),
]
)
# Should ask ML to predict whether constraint should be removed # Should ask ML to predict whether constraint should be removed
component.classifiers["type-a"].predict_proba.assert_called_once_with([[1.0, 0.0], [0.5, 0.5]]) component.classifiers["type-a"].predict_proba.assert_called_once_with(
[[1.0, 0.0], [0.5, 0.5]]
)
component.classifiers["type-b"].predict_proba.assert_called_once_with([[1.0]]) component.classifiers["type-b"].predict_proba.assert_called_once_with([[1.0]])
# Should ask internal solver to remove constraints predicted as redundant # Should ask internal solver to remove constraints predicted as redundant
assert internal.extract_constraint.call_count == 2 assert internal.extract_constraint.call_count == 2
internal.extract_constraint.assert_has_calls([ internal.extract_constraint.assert_has_calls(
call("c3"), call("c4"), [
]) call("c3"),
call("c4"),
]
)
# LearningSolver calls after_solve # LearningSolver calls after_solve
component.after_solve(solver, instance, None, None) component.after_solve(solver, instance, None, None)
@@ -111,8 +130,7 @@ def test_usage():
def test_usage_with_check_dropped(): def test_usage_with_check_dropped():
solver, internal, instance, classifiers = _setup() solver, internal, instance, classifiers = _setup()
component = RelaxationComponent(check_dropped=True, component = RelaxationComponent(check_dropped=True, violation_tolerance=1e-3)
violation_tolerance=1e-3)
component.classifiers = classifiers component.classifiers = classifiers
# LearningSolver call before_solve # LearningSolver call before_solve
@@ -120,9 +138,12 @@ def test_usage_with_check_dropped():
# Assert constraints are extracted # Assert constraints are extracted
assert internal.extract_constraint.call_count == 2 assert internal.extract_constraint.call_count == 2
internal.extract_constraint.assert_has_calls([ internal.extract_constraint.assert_has_calls(
call("c3"), call("c4"), [
]) call("c3"),
call("c4"),
]
)
# LearningSolver calls iteration_cb (first time) # LearningSolver calls iteration_cb (first time)
should_repeat = component.iteration_cb(solver, instance, None) should_repeat = component.iteration_cb(solver, instance, None)
@@ -131,15 +152,15 @@ def test_usage_with_check_dropped():
assert should_repeat assert should_repeat
# Should ask solver if removed constraints are satisfied (mock always returns false) # Should ask solver if removed constraints are satisfied (mock always returns false)
internal.is_constraint_satisfied.assert_has_calls([ internal.is_constraint_satisfied.assert_has_calls(
call("<c3>", 1e-3), [
call("<c4>", 1e-3), call("<c3>", 1e-3),
]) call("<c4>", 1e-3),
]
)
# Should add constraints back to LP relaxation # Should add constraints back to LP relaxation
internal.add_constraint.assert_has_calls([ internal.add_constraint.assert_has_calls([call("<c3>"), call("<c4>")])
call("<c3>"), call("<c4>")
])
# LearningSolver calls iteration_cb (second time) # LearningSolver calls iteration_cb (second time)
should_repeat = component.iteration_cb(solver, instance, None) should_repeat = component.iteration_cb(solver, instance, None)
@@ -148,21 +169,22 @@ def test_usage_with_check_dropped():
def test_x_y_fit_predict_evaluate(): def test_x_y_fit_predict_evaluate():
instances = [Mock(spec=Instance), Mock(spec=Instance)] instances = [Mock(spec=Instance), Mock(spec=Instance)]
component = RelaxationComponent(slack_tolerance=0.05, component = RelaxationComponent(slack_tolerance=0.05, threshold=0.80)
threshold=0.80)
component.classifiers = { component.classifiers = {
"type-a": Mock(spec=Classifier), "type-a": Mock(spec=Classifier),
"type-b": Mock(spec=Classifier), "type-b": Mock(spec=Classifier),
} }
component.classifiers["type-a"].predict_proba = \ component.classifiers["type-a"].predict_proba = Mock(
Mock(return_value=[ return_value=[
[0.20, 0.80], [0.20, 0.80],
]) ]
component.classifiers["type-b"].predict_proba = \ )
Mock(return_value=[ component.classifiers["type-b"].predict_proba = Mock(
return_value=[
[0.50, 0.50], [0.50, 0.50],
[0.05, 0.95], [0.05, 0.95],
]) ]
)
# First mock instance # First mock instance
instances[0].slacks = { instances[0].slacks = {
@@ -171,17 +193,21 @@ def test_x_y_fit_predict_evaluate():
"c3": 0.00, "c3": 0.00,
"c4": 30.0, "c4": 30.0,
} }
instances[0].get_constraint_category = Mock(side_effect=lambda cid: { instances[0].get_constraint_category = Mock(
"c1": None, side_effect=lambda cid: {
"c2": "type-a", "c1": None,
"c3": "type-a", "c2": "type-a",
"c4": "type-b", "c3": "type-a",
}[cid]) "c4": "type-b",
instances[0].get_constraint_features = Mock(side_effect=lambda cid: { }[cid]
"c2": [1.0, 0.0], )
"c3": [0.5, 0.5], instances[0].get_constraint_features = Mock(
"c4": [1.0], side_effect=lambda cid: {
}[cid]) "c2": [1.0, 0.0],
"c3": [0.5, 0.5],
"c4": [1.0],
}[cid]
)
# Second mock instance # Second mock instance
instances[1].slacks = { instances[1].slacks = {
@@ -190,26 +216,27 @@ def test_x_y_fit_predict_evaluate():
"c4": 0.00, "c4": 0.00,
"c5": 0.00, "c5": 0.00,
} }
instances[1].get_constraint_category = Mock(side_effect=lambda cid: { instances[1].get_constraint_category = Mock(
"c1": None, side_effect=lambda cid: {
"c3": "type-a", "c1": None,
"c4": "type-b", "c3": "type-a",
"c5": "type-b", "c4": "type-b",
}[cid]) "c5": "type-b",
instances[1].get_constraint_features = Mock(side_effect=lambda cid: { }[cid]
"c3": [0.3, 0.4], )
"c4": [0.7], instances[1].get_constraint_features = Mock(
"c5": [0.8], side_effect=lambda cid: {
}[cid]) "c3": [0.3, 0.4],
"c4": [0.7],
"c5": [0.8],
}[cid]
)
expected_x = { expected_x = {
"type-a": [[1.0, 0.0], [0.5, 0.5], [0.3, 0.4]], "type-a": [[1.0, 0.0], [0.5, 0.5], [0.3, 0.4]],
"type-b": [[1.0], [0.7], [0.8]], "type-b": [[1.0], [0.7], [0.8]],
} }
expected_y = { expected_y = {"type-a": [[0], [0], [1]], "type-b": [[1], [0], [0]]}
"type-a": [[0], [0], [1]],
"type-b": [[1], [0], [0]]
}
# Should build X and Y matrices correctly # Should build X and Y matrices correctly
assert component.x(instances) == expected_x assert component.x(instances) == expected_x
@@ -217,13 +244,16 @@ def test_x_y_fit_predict_evaluate():
# Should pass along X and Y matrices to classifiers # Should pass along X and Y matrices to classifiers
component.fit(instances) component.fit(instances)
component.classifiers["type-a"].fit.assert_called_with(expected_x["type-a"], expected_y["type-a"]) component.classifiers["type-a"].fit.assert_called_with(
component.classifiers["type-b"].fit.assert_called_with(expected_x["type-b"], expected_y["type-b"]) expected_x["type-a"],
expected_y["type-a"],
)
component.classifiers["type-b"].fit.assert_called_with(
expected_x["type-b"],
expected_y["type-b"],
)
assert component.predict(expected_x) == { assert component.predict(expected_x) == {"type-a": [[1]], "type-b": [[0], [1]]}
"type-a": [[1]],
"type-b": [[0], [1]]
}
ev = component.evaluate(instances[1]) ev = component.evaluate(instances[1])
assert ev["True positive"] == 1 assert ev["True positive"] == 1

79
miplearn/extractors.py
View File

@@ -18,10 +18,10 @@ class InstanceIterator:
def __init__(self, instances): def __init__(self, instances):
self.instances = instances self.instances = instances
self.current = 0 self.current = 0
def __iter__(self): def __iter__(self):
return self return self
def __next__(self): def __next__(self):
if self.current >= len(self.instances): if self.current >= len(self.instances):
raise StopIteration raise StopIteration
@@ -40,9 +40,9 @@ class InstanceIterator:
class Extractor(ABC): class Extractor(ABC):
@abstractmethod @abstractmethod
def extract(self, instances,): def extract(self, instances):
pass pass
@staticmethod @staticmethod
def split_variables(instance): def split_variables(instance):
assert hasattr(instance, "lp_solution") assert hasattr(instance, "lp_solution")
@@ -57,13 +57,15 @@ class Extractor(ABC):
result[category] += [(var_name, index)] result[category] += [(var_name, index)]
return result return result
class VariableFeaturesExtractor(Extractor): class VariableFeaturesExtractor(Extractor):
def extract(self, instances): def extract(self, instances):
result = {} result = {}
for instance in tqdm(InstanceIterator(instances), for instance in tqdm(
desc="Extract (vars)", InstanceIterator(instances),
disable=len(instances) < 5): desc="Extract (vars)",
disable=len(instances) < 5,
):
instance_features = instance.get_instance_features() instance_features = instance.get_instance_features()
var_split = self.split_variables(instance) var_split = self.split_variables(instance)
for (category, var_index_pairs) in var_split.items(): for (category, var_index_pairs) in var_split.items():
@@ -71,9 +73,9 @@ class VariableFeaturesExtractor(Extractor):
result[category] = [] result[category] = []
for (var_name, index) in var_index_pairs: for (var_name, index) in var_index_pairs:
result[category] += [ result[category] += [
instance_features.tolist() + \ instance_features.tolist()
instance.get_variable_features(var_name, index).tolist() + \ + instance.get_variable_features(var_name, index).tolist()
[instance.lp_solution[var_name][index]] + [instance.lp_solution[var_name][index]]
] ]
for category in result: for category in result:
result[category] = np.array(result[category]) result[category] = np.array(result[category])
@@ -83,12 +85,14 @@ class VariableFeaturesExtractor(Extractor):
class SolutionExtractor(Extractor): class SolutionExtractor(Extractor):
def __init__(self, relaxation=False): def __init__(self, relaxation=False):
self.relaxation = relaxation self.relaxation = relaxation
def extract(self, instances): def extract(self, instances):
result = {} result = {}
for instance in tqdm(InstanceIterator(instances), for instance in tqdm(
desc="Extract (solution)", InstanceIterator(instances),
disable=len(instances) < 5): desc="Extract (solution)",
disable=len(instances) < 5,
):
var_split = self.split_variables(instance) var_split = self.split_variables(instance)
for (category, var_index_pairs) in var_split.items(): for (category, var_index_pairs) in var_split.items():
if category not in result: if category not in result:
@@ -103,33 +107,40 @@ class SolutionExtractor(Extractor):
else: else:
result[category] += [[1 - v, v]] result[category] += [[1 - v, v]]
for category in result: for category in result:
result[category] = np.array(result[category]) result[category] = np.array(result[category])
return result return result
class InstanceFeaturesExtractor(Extractor): class InstanceFeaturesExtractor(Extractor):
def extract(self, instances): def extract(self, instances):
return np.vstack([ return np.vstack(
np.hstack([ [
instance.get_instance_features(), np.hstack(
instance.lp_value, [
]) instance.get_instance_features(),
for instance in InstanceIterator(instances) instance.lp_value,
]) ]
)
for instance in InstanceIterator(instances)
]
)
class ObjectiveValueExtractor(Extractor): class ObjectiveValueExtractor(Extractor):
def __init__(self, kind="lp"): def __init__(self, kind="lp"):
assert kind in ["lower bound", "upper bound", "lp"] assert kind in ["lower bound", "upper bound", "lp"]
self.kind = kind self.kind = kind
def extract(self, instances): def extract(self, instances):
if self.kind == "lower bound": if self.kind == "lower bound":
return np.array([[instance.lower_bound] return np.array(
for instance in InstanceIterator(instances)]) [[instance.lower_bound] for instance in InstanceIterator(instances)]
)
if self.kind == "upper bound": if self.kind == "upper bound":
return np.array([[instance.upper_bound] return np.array(
for instance in InstanceIterator(instances)]) [[instance.upper_bound] for instance in InstanceIterator(instances)]
)
if self.kind == "lp": if self.kind == "lp":
return np.array([[instance.lp_value] return np.array(
for instance in InstanceIterator(instances)]) [[instance.lp_value] for instance in InstanceIterator(instances)]
)

36
miplearn/instance.py
View File

@@ -12,7 +12,7 @@ import numpy as np
class Instance(ABC): class Instance(ABC):
""" """
Abstract class holding all the data necessary to generate a concrete model of the problem. Abstract class holding all the data necessary to generate a concrete model of the problem.
In the knapsack problem, for example, this class could hold the number of items, their weights In the knapsack problem, for example, this class could hold the number of items, their weights
and costs, as well as the size of the knapsack. Objects implementing this class are able to and costs, as well as the size of the knapsack. Objects implementing this class are able to
convert themselves into a concrete optimization model, which can be optimized by a solver, or convert themselves into a concrete optimization model, which can be optimized by a solver, or
@@ -29,17 +29,17 @@ class Instance(ABC):
def get_instance_features(self): def get_instance_features(self):
""" """
Returns a 1-dimensional Numpy array of (numerical) features describing the entire instance. Returns a 1-dimensional Numpy array of (numerical) features describing the entire instance.
The array is used by LearningSolver to determine how similar two instances are. It may also The array is used by LearningSolver to determine how similar two instances are. It may also
be used to predict, in combination with variable-specific features, the values of binary be used to predict, in combination with variable-specific features, the values of binary
decision variables in the problem. decision variables in the problem.
There is not necessarily a one-to-one correspondence between models and instance features: There is not necessarily a one-to-one correspondence between models and instance features:
the features may encode only part of the data necessary to generate the complete model. the features may encode only part of the data necessary to generate the complete model.
Features may also be statistics computed from the original data. For example, in the Features may also be statistics computed from the original data. For example, in the
knapsack problem, an implementation may decide to provide as instance features only knapsack problem, an implementation may decide to provide as instance features only
the average weights, average prices, number of items and the size of the knapsack. the average weights, average prices, number of items and the size of the knapsack.
The returned array MUST have the same length for all relevant instances of the problem. If The returned array MUST have the same length for all relevant instances of the problem. If
two instances map into arrays of different lengths, they cannot be solved by the same two instances map into arrays of different lengths, they cannot be solved by the same
LearningSolver object. LearningSolver object.
@@ -52,16 +52,16 @@ class Instance(ABC):
""" """
Returns a 1-dimensional array of (numerical) features describing a particular decision Returns a 1-dimensional array of (numerical) features describing a particular decision
variable. variable.
The argument `var` is a pyomo.core.Var object, which represents a collection of decision The argument `var` is a pyomo.core.Var object, which represents a collection of decision
variables. The argument `index` specifies which variable in the collection is the relevant variables. The argument `index` specifies which variable in the collection is the relevant
one. one.
In combination with instance features, variable features are used by LearningSolver to In combination with instance features, variable features are used by LearningSolver to
predict, among other things, the optimal value of each decision variable before the predict, among other things, the optimal value of each decision variable before the
optimization takes place. In the knapsack problem, for example, an implementation could optimization takes place. In the knapsack problem, for example, an implementation could
provide as variable features the weight and the price of a specific item. provide as variable features the weight and the price of a specific item.
Like instance features, the arrays returned by this method MUST have the same length for Like instance features, the arrays returned by this method MUST have the same length for
all variables within the same category, for all relevant instances of the problem. all variables within the same category, for all relevant instances of the problem.
@@ -73,7 +73,7 @@ class Instance(ABC):
""" """
Returns the category (a string, an integer or any hashable type) for each decision Returns the category (a string, an integer or any hashable type) for each decision
variable. variable.
If two variables have the same category, LearningSolver will use the same internal ML If two variables have the same category, LearningSolver will use the same internal ML
model to predict the values of both variables. If the returned category is None, ML model to predict the values of both variables. If the returned category is None, ML
models will ignore the variable. models will ignore the variable.
@@ -100,18 +100,18 @@ class Instance(ABC):
def find_violated_lazy_constraints(self, model): def find_violated_lazy_constraints(self, model):
""" """
Returns lazy constraint violations found for the current solution. Returns lazy constraint violations found for the current solution.
After solving a model, LearningSolver will ask the instance to identify which lazy After solving a model, LearningSolver will ask the instance to identify which lazy
constraints are violated by the current solution. For each identified violation, constraints are violated by the current solution. For each identified violation,
LearningSolver will then call the build_lazy_constraint, add the generated Pyomo LearningSolver will then call the build_lazy_constraint, add the generated Pyomo
constraint to the model, then resolve the problem. The process repeats until no further constraint to the model, then resolve the problem. The process repeats until no further
lazy constraint violations are found. lazy constraint violations are found.
Each "violation" is simply a string, a tuple or any other hashable type which allows the Each "violation" is simply a string, a tuple or any other hashable type which allows the
instance to identify unambiguously which lazy constraint should be generated. In the instance to identify unambiguously which lazy constraint should be generated. In the
Traveling Salesman Problem, for example, a subtour violation could be a frozen set Traveling Salesman Problem, for example, a subtour violation could be a frozen set
containing the cities in the subtour. containing the cities in the subtour.
For a concrete example, see TravelingSalesmanInstance. For a concrete example, see TravelingSalesmanInstance.
""" """
return [] return []
@@ -119,17 +119,17 @@ class Instance(ABC):
def build_lazy_constraint(self, model, violation): def build_lazy_constraint(self, model, violation):
""" """
Returns a Pyomo constraint which fixes a given violation. Returns a Pyomo constraint which fixes a given violation.
This method is typically called immediately after find_violated_lazy_constraints. The violation object This method is typically called immediately after find_violated_lazy_constraints. The violation object
provided to this method is exactly the same object returned earlier by find_violated_lazy_constraints. provided to this method is exactly the same object returned earlier by find_violated_lazy_constraints.
After some training, LearningSolver may decide to proactively build some lazy constraints After some training, LearningSolver may decide to proactively build some lazy constraints
at the beginning of the optimization process, before a solution is even available. In this at the beginning of the optimization process, before a solution is even available. In this
case, build_lazy_constraints will be called without a corresponding call to case, build_lazy_constraints will be called without a corresponding call to
find_violated_lazy_constraints. find_violated_lazy_constraints.
The implementation should not directly add the constraint to the model. The constraint The implementation should not directly add the constraint to the model. The constraint
will be added by LearningSolver after the method returns. will be added by LearningSolver after the method returns.
For a concrete example, see TravelingSalesmanInstance. For a concrete example, see TravelingSalesmanInstance.
""" """
pass pass
@@ -141,11 +141,11 @@ class Instance(ABC):
pass pass
def load(self, filename): def load(self, filename):
with gzip.GzipFile(filename, 'r') as f: with gzip.GzipFile(filename, "r") as f:
data = json.loads(f.read().decode('utf-8')) data = json.loads(f.read().decode("utf-8"))
self.__dict__ = data self.__dict__ = data
def dump(self, filename): def dump(self, filename):
data = json.dumps(self.__dict__, indent=2).encode('utf-8') data = json.dumps(self.__dict__, indent=2).encode("utf-8")
with gzip.GzipFile(filename, 'w') as f: with gzip.GzipFile(filename, "w") as f:
f.write(data) f.write(data)

27
miplearn/log.py
View File

@@ -7,7 +7,8 @@ import logging
import time import time
import sys import sys
class TimeFormatter():
class TimeFormatter:
def __init__(self, start_time, log_colors): def __init__(self, start_time, log_colors):
self.start_time = start_time self.start_time = start_time
self.log_colors = log_colors self.log_colors = log_colors
@@ -19,21 +20,23 @@ class TimeFormatter():
color = self.log_colors["yellow"] color = self.log_colors["yellow"]
else: else:
color = self.log_colors["green"] color = self.log_colors["green"]
return "%s[%12.3f]%s %s" % (color, return "%s[%12.3f]%s %s" % (
record.created - self.start_time, color,
self.log_colors["reset"], record.created - self.start_time,
record.getMessage()) self.log_colors["reset"],
record.getMessage(),
)
def setup_logger(start_time=None,
force_color=False): def setup_logger(start_time=None, force_color=False):
if start_time is None: if start_time is None:
start_time = time.time() start_time = time.time()
if sys.stdout.isatty() or force_color: if sys.stdout.isatty() or force_color:
log_colors = { log_colors = {
"green": '\033[92m', "green": "\033[92m",
"yellow": '\033[93m', "yellow": "\033[93m",
"red": '\033[91m', "red": "\033[91m",
"reset": '\033[0m', "reset": "\033[0m",
} }
else: else:
log_colors = { log_colors = {
@@ -41,7 +44,7 @@ def setup_logger(start_time=None,
"yellow": "", "yellow": "",
"red": "", "red": "",
"reset": "", "reset": "",
} }
handler = logging.StreamHandler() handler = logging.StreamHandler()
handler.setFormatter(TimeFormatter(start_time, log_colors)) handler.setFormatter(TimeFormatter(start_time, log_colors))
logging.getLogger().addHandler(handler) logging.getLogger().addHandler(handler)

190
miplearn/problems/knapsack.py
View File

@@ -17,44 +17,45 @@ class ChallengeA:
- K = 500, u ~ U(0., 1.) - K = 500, u ~ U(0., 1.)
- alpha = 0.25 - alpha = 0.25
""" """
def __init__(self,
seed=42, def __init__(
n_training_instances=500, self,
n_test_instances=50): seed=42,
n_training_instances=500,
n_test_instances=50,
):
np.random.seed(seed) np.random.seed(seed)
self.gen = MultiKnapsackGenerator(n=randint(low=250, high=251), self.gen = MultiKnapsackGenerator(
m=randint(low=10, high=11), n=randint(low=250, high=251),
w=uniform(loc=0.0, scale=1000.0), m=randint(low=10, high=11),
K=uniform(loc=500.0, scale=0.0), w=uniform(loc=0.0, scale=1000.0),
u=uniform(loc=0.0, scale=1.0), K=uniform(loc=500.0, scale=0.0),
alpha=uniform(loc=0.25, scale=0.0), u=uniform(loc=0.0, scale=1.0),
fix_w=True, alpha=uniform(loc=0.25, scale=0.0),
w_jitter=uniform(loc=0.95, scale=0.1), fix_w=True,
) w_jitter=uniform(loc=0.95, scale=0.1),
)
np.random.seed(seed + 1) np.random.seed(seed + 1)
self.training_instances = self.gen.generate(n_training_instances) self.training_instances = self.gen.generate(n_training_instances)
np.random.seed(seed + 2) np.random.seed(seed + 2)
self.test_instances = self.gen.generate(n_test_instances) self.test_instances = self.gen.generate(n_test_instances)
class MultiKnapsackInstance(Instance): class MultiKnapsackInstance(Instance):
"""Representation of the Multidimensional 0-1 Knapsack Problem. """Representation of the Multidimensional 0-1 Knapsack Problem.
Given a set of n items and m knapsacks, the problem is to find a subset of items S maximizing Given a set of n items and m knapsacks, the problem is to find a subset of items S maximizing
sum(prices[i] for i in S). If selected, each item i occupies weights[i,j] units of space in sum(prices[i] for i in S). If selected, each item i occupies weights[i,j] units of space in
each knapsack j. Furthermore, each knapsack j has limited storage space, given by capacities[j]. each knapsack j. Furthermore, each knapsack j has limited storage space, given by capacities[j].
This implementation assigns a different category for each decision variable, and therefore This implementation assigns a different category for each decision variable, and therefore
trains one ML model per variable. It is only suitable when training and test instances have trains one ML model per variable. It is only suitable when training and test instances have
same size and items don't shuffle around. same size and items don't shuffle around.
""" """
def __init__(self, def __init__(self, prices, capacities, weights):
prices,
capacities,
weights):
assert isinstance(prices, np.ndarray) assert isinstance(prices, np.ndarray)
assert isinstance(capacities, np.ndarray) assert isinstance(capacities, np.ndarray)
assert isinstance(weights, np.ndarray) assert isinstance(weights, np.ndarray)
@@ -65,83 +66,92 @@ class MultiKnapsackInstance(Instance):
self.prices = prices self.prices = prices
self.capacities = capacities self.capacities = capacities
self.weights = weights self.weights = weights
def to_model(self): def to_model(self):
model = pe.ConcreteModel() model = pe.ConcreteModel()
model.x = pe.Var(range(self.n), domain=pe.Binary) model.x = pe.Var(range(self.n), domain=pe.Binary)
model.OBJ = pe.Objective(rule=lambda model: sum(model.x[j] * self.prices[j] model.OBJ = pe.Objective(
for j in range(self.n)), rule=lambda model: sum(model.x[j] * self.prices[j] for j in range(self.n)),
sense=pe.maximize) sense=pe.maximize,
)
model.eq_capacity = pe.ConstraintList() model.eq_capacity = pe.ConstraintList()
for i in range(self.m): for i in range(self.m):
model.eq_capacity.add(sum(model.x[j] * self.weights[i,j] model.eq_capacity.add(
for j in range(self.n)) <= self.capacities[i]) sum(model.x[j] * self.weights[i, j] for j in range(self.n))
<= self.capacities[i]
)
return model return model
def get_instance_features(self): def get_instance_features(self):
return np.hstack([ return np.hstack(
np.mean(self.prices), [
self.capacities, np.mean(self.prices),
]) self.capacities,
]
)
def get_variable_features(self, var, index): def get_variable_features(self, var, index):
return np.hstack([ return np.hstack(
self.prices[index], [
self.weights[:, index], self.prices[index],
]) self.weights[:, index],
]
)
# def get_variable_category(self, var, index): # def get_variable_category(self, var, index):
# return index # return index
class MultiKnapsackGenerator: class MultiKnapsackGenerator:
def __init__(self, def __init__(
n=randint(low=100, high=101), self,
m=randint(low=30, high=31), n=randint(low=100, high=101),
w=randint(low=0, high=1000), m=randint(low=30, high=31),
K=randint(low=500, high=500), w=randint(low=0, high=1000),
u=uniform(loc=0.0, scale=1.0), K=randint(low=500, high=500),
alpha=uniform(loc=0.25, scale=0.0), u=uniform(loc=0.0, scale=1.0),
fix_w=False, alpha=uniform(loc=0.25, scale=0.0),
w_jitter=uniform(loc=1.0, scale=0.0), fix_w=False,
round=True, w_jitter=uniform(loc=1.0, scale=0.0),
): round=True,
):
"""Initialize the problem generator. """Initialize the problem generator.
Instances have a random number of items (or variables) and a random number of knapsacks Instances have a random number of items (or variables) and a random number of knapsacks
(or constraints), as specified by the provided probability distributions `n` and `m`, (or constraints), as specified by the provided probability distributions `n` and `m`,
respectively. The weight of each item `i` on knapsack `j` is sampled independently from respectively. The weight of each item `i` on knapsack `j` is sampled independently from
the provided distribution `w`. The capacity of knapsack `j` is set to: the provided distribution `w`. The capacity of knapsack `j` is set to:
alpha_j * sum(w[i,j] for i in range(n)), alpha_j * sum(w[i,j] for i in range(n)),
where `alpha_j`, the tightness ratio, is sampled from the provided probability where `alpha_j`, the tightness ratio, is sampled from the provided probability
distribution `alpha`. To make the instances more challenging, the costs of the items distribution `alpha`. To make the instances more challenging, the costs of the items
are linearly correlated to their average weights. More specifically, the weight of each are linearly correlated to their average weights. More specifically, the weight of each
item `i` is set to: item `i` is set to:
sum(w[i,j]/m for j in range(m)) + K * u_i, sum(w[i,j]/m for j in range(m)) + K * u_i,
where `K`, the correlation coefficient, and `u_i`, the correlation multiplier, are sampled where `K`, the correlation coefficient, and `u_i`, the correlation multiplier, are sampled
from the provided probability distributions. Note that `K` is only sample once for the from the provided probability distributions. Note that `K` is only sample once for the
entire instance. entire instance.
If fix_w=True is provided, then w[i,j] are kept the same in all generated instances. This If fix_w=True is provided, then w[i,j] are kept the same in all generated instances. This
also implies that n and m are kept fixed. Although the prices and capacities are derived also implies that n and m are kept fixed. Although the prices and capacities are derived
from w[i,j], as long as u and K are not constants, the generated instances will still not from w[i,j], as long as u and K are not constants, the generated instances will still not
be completely identical. be completely identical.
If a probability distribution w_jitter is provided, then item weights will be set to If a probability distribution w_jitter is provided, then item weights will be set to
w[i,j] * gamma[i,j] where gamma[i,j] is sampled from w_jitter. When combined with w[i,j] * gamma[i,j] where gamma[i,j] is sampled from w_jitter. When combined with
fix_w=True, this argument may be used to generate instances where the weight of each item fix_w=True, this argument may be used to generate instances where the weight of each item
is roughly the same, but not exactly identical, across all instances. The prices of the is roughly the same, but not exactly identical, across all instances. The prices of the
items and the capacities of the knapsacks will be calculated as above, but using these items and the capacities of the knapsacks will be calculated as above, but using these
perturbed weights instead. perturbed weights instead.
By default, all generated prices, weights and capacities are rounded to the nearest integer By default, all generated prices, weights and capacities are rounded to the nearest integer
number. If `round=False` is provided, this rounding will be disabled. number. If `round=False` is provided, this rounding will be disabled.
Parameters Parameters
---------- ----------
n: rv_discrete n: rv_discrete
@@ -168,11 +178,14 @@ class MultiKnapsackGenerator:
assert isinstance(w, rv_frozen), "w should be a SciPy probability distribution" assert isinstance(w, rv_frozen), "w should be a SciPy probability distribution"
assert isinstance(K, rv_frozen), "K should be a SciPy probability distribution" assert isinstance(K, rv_frozen), "K should be a SciPy probability distribution"
assert isinstance(u, rv_frozen), "u should be a SciPy probability distribution" assert isinstance(u, rv_frozen), "u should be a SciPy probability distribution"
assert isinstance(alpha, rv_frozen), "alpha should be a SciPy probability distribution" assert isinstance(
alpha, rv_frozen
), "alpha should be a SciPy probability distribution"
assert isinstance(fix_w, bool), "fix_w should be boolean" assert isinstance(fix_w, bool), "fix_w should be boolean"
assert isinstance(w_jitter, rv_frozen), \ assert isinstance(
"w_jitter should be a SciPy probability distribution" w_jitter, rv_frozen
), "w_jitter should be a SciPy probability distribution"
self.n = n self.n = n
self.m = m self.m = m
self.w = w self.w = w
@@ -181,7 +194,7 @@ class MultiKnapsackGenerator:
self.alpha = alpha self.alpha = alpha
self.w_jitter = w_jitter self.w_jitter = w_jitter
self.round = round self.round = round
if fix_w: if fix_w:
self.fix_n = self.n.rvs() self.fix_n = self.n.rvs()
self.fix_m = self.m.rvs() self.fix_m = self.m.rvs()
@@ -194,7 +207,7 @@ class MultiKnapsackGenerator:
self.fix_w = None self.fix_w = None
self.fix_u = None self.fix_u = None
self.fix_K = None self.fix_K = None
def generate(self, n_samples): def generate(self, n_samples):
def _sample(): def _sample():
if self.fix_w is not None: if self.fix_w is not None:
@@ -211,20 +224,22 @@ class MultiKnapsackGenerator:
K = self.K.rvs() K = self.K.rvs()
w = w * np.array([self.w_jitter.rvs(n) for _ in range(m)]) w = w * np.array([self.w_jitter.rvs(n) for _ in range(m)])
alpha = self.alpha.rvs(m) alpha = self.alpha.rvs(m)
p = np.array([w[:,j].sum() / m + K * u[j] for j in range(n)]) p = np.array([w[:, j].sum() / m + K * u[j] for j in range(n)])
b = np.array([w[i,:].sum() * alpha[i] for i in range(m)]) b = np.array([w[i, :].sum() * alpha[i] for i in range(m)])
if self.round: if self.round:
p = p.round() p = p.round()
b = b.round() b = b.round()
w = w.round() w = w.round()
return MultiKnapsackInstance(p, b, w) return MultiKnapsackInstance(p, b, w)
return [_sample() for _ in range(n_samples)] return [_sample() for _ in range(n_samples)]
class KnapsackInstance(Instance): class KnapsackInstance(Instance):
""" """
Simpler (one-dimensional) Knapsack Problem, used for testing. Simpler (one-dimensional) Knapsack Problem, used for testing.
""" """
def __init__(self, weights, prices, capacity): def __init__(self, weights, prices, capacity):
self.weights = weights self.weights = weights
self.prices = prices self.prices = prices
@@ -234,23 +249,29 @@ class KnapsackInstance(Instance):
model = pe.ConcreteModel() model = pe.ConcreteModel()
items = range(len(self.weights)) items = range(len(self.weights))
model.x = pe.Var(items, domain=pe.Binary) model.x = pe.Var(items, domain=pe.Binary)
model.OBJ = pe.Objective(expr=sum(model.x[v] * self.prices[v] for v in items), model.OBJ = pe.Objective(
sense=pe.maximize) expr=sum(model.x[v] * self.prices[v] for v in items), sense=pe.maximize
model.eq_capacity = pe.Constraint(expr=sum(model.x[v] * self.weights[v] )
for v in items) <= self.capacity) model.eq_capacity = pe.Constraint(
expr=sum(model.x[v] * self.weights[v] for v in items) <= self.capacity
)
return model return model
def get_instance_features(self): def get_instance_features(self):
return np.array([ return np.array(
self.capacity, [
np.average(self.weights), self.capacity,
]) np.average(self.weights),
]
)
def get_variable_features(self, var, index): def get_variable_features(self, var, index):
return np.array([ return np.array(
self.weights[index], [
self.prices[index], self.weights[index],
]) self.prices[index],
]
)
class GurobiKnapsackInstance(KnapsackInstance): class GurobiKnapsackInstance(KnapsackInstance):
@@ -258,6 +279,7 @@ class GurobiKnapsackInstance(KnapsackInstance):
Simpler (one-dimensional) knapsack instance, implemented directly in Gurobi Simpler (one-dimensional) knapsack instance, implemented directly in Gurobi
instead of Pyomo, used for testing. instead of Pyomo, used for testing.
""" """
def __init__(self, weights, prices, capacity): def __init__(self, weights, prices, capacity):
super().__init__(weights, prices, capacity) super().__init__(weights, prices, capacity)
@@ -268,9 +290,11 @@ class GurobiKnapsackInstance(KnapsackInstance):
model = gp.Model("Knapsack") model = gp.Model("Knapsack")
n = len(self.weights) n = len(self.weights)
x = model.addVars(n, vtype=GRB.BINARY, name="x") x = model.addVars(n, vtype=GRB.BINARY, name="x")
model.addConstr(gp.quicksum(x[i] * self.weights[i] model.addConstr(
for i in range(n)) <= self.capacity, gp.quicksum(x[i] * self.weights[i] for i in range(n)) <= self.capacity,
"eq_capacity") "eq_capacity",
model.setObjective(gp.quicksum(x[i] * self.prices[i] )
for i in range(n)), GRB.MAXIMIZE) model.setObjective(
gp.quicksum(x[i] * self.prices[i] for i in range(n)), GRB.MAXIMIZE
)
return model return model

65
miplearn/problems/stab.py
View File

@@ -12,44 +12,49 @@ from scipy.stats.distributions import rv_frozen
class ChallengeA: class ChallengeA:
def __init__(self, def __init__(
seed=42, self,
n_training_instances=500, seed=42,
n_test_instances=50, n_training_instances=500,
): n_test_instances=50,
):
np.random.seed(seed) np.random.seed(seed)
self.generator = MaxWeightStableSetGenerator(w=uniform(loc=100., scale=50.), self.generator = MaxWeightStableSetGenerator(
n=randint(low=200, high=201), w=uniform(loc=100.0, scale=50.0),
p=uniform(loc=0.05, scale=0.0), n=randint(low=200, high=201),
fix_graph=True) p=uniform(loc=0.05, scale=0.0),
fix_graph=True,
)
np.random.seed(seed + 1) np.random.seed(seed + 1)
self.training_instances = self.generator.generate(n_training_instances) self.training_instances = self.generator.generate(n_training_instances)
np.random.seed(seed + 2) np.random.seed(seed + 2)
self.test_instances = self.generator.generate(n_test_instances) self.test_instances = self.generator.generate(n_test_instances)
class MaxWeightStableSetGenerator: class MaxWeightStableSetGenerator:
"""Random instance generator for the Maximum-Weight Stable Set Problem. """Random instance generator for the Maximum-Weight Stable Set Problem.
The generator has two modes of operation. When `fix_graph=True` is provided, one random The generator has two modes of operation. When `fix_graph=True` is provided, one random
Erdős-Rényi graph $G_{n,p}$ is generated in the constructor, where $n$ and $p$ are sampled Erdős-Rényi graph $G_{n,p}$ is generated in the constructor, where $n$ and $p$ are sampled
from user-provided probability distributions `n` and `p`. To generate each instance, the from user-provided probability distributions `n` and `p`. To generate each instance, the
generator independently samples each $w_v$ from the user-provided probability distribution `w`. generator independently samples each $w_v$ from the user-provided probability distribution `w`.
When `fix_graph=False`, a new random graph is generated for each instance; the remaining When `fix_graph=False`, a new random graph is generated for each instance; the remaining
parameters are sampled in the same way. parameters are sampled in the same way.
""" """
def __init__(self, def __init__(
w=uniform(loc=10.0, scale=1.0), self,
n=randint(low=250, high=251), w=uniform(loc=10.0, scale=1.0),
p=uniform(loc=0.05, scale=0.0), n=randint(low=250, high=251),
fix_graph=True): p=uniform(loc=0.05, scale=0.0),
fix_graph=True,
):
"""Initialize the problem generator. """Initialize the problem generator.
Parameters Parameters
---------- ----------
w: rv_continuous w: rv_continuous
@@ -69,7 +74,7 @@ class MaxWeightStableSetGenerator:
self.graph = None self.graph = None
if fix_graph: if fix_graph:
self.graph = self._generate_graph() self.graph = self._generate_graph()
def generate(self, n_samples): def generate(self, n_samples):
def _sample(): def _sample():
if self.graph is not None: if self.graph is not None:
@@ -78,22 +83,23 @@ class MaxWeightStableSetGenerator:
graph = self._generate_graph() graph = self._generate_graph()
weights = self.w.rvs(graph.number_of_nodes()) weights = self.w.rvs(graph.number_of_nodes())
return MaxWeightStableSetInstance(graph, weights) return MaxWeightStableSetInstance(graph, weights)
return [_sample() for _ in range(n_samples)] return [_sample() for _ in range(n_samples)]
def _generate_graph(self): def _generate_graph(self):
return nx.generators.random_graphs.binomial_graph(self.n.rvs(), self.p.rvs()) return nx.generators.random_graphs.binomial_graph(self.n.rvs(), self.p.rvs())
class MaxWeightStableSetInstance(Instance): class MaxWeightStableSetInstance(Instance):
"""An instance of the Maximum-Weight Stable Set Problem. """An instance of the Maximum-Weight Stable Set Problem.
Given a graph G=(V,E) and a weight w_v for each vertex v, the problem asks for a stable Given a graph G=(V,E) and a weight w_v for each vertex v, the problem asks for a stable
set S of G maximizing sum(w_v for v in S). A stable set (also called independent set) is set S of G maximizing sum(w_v for v in S). A stable set (also called independent set) is
a subset of vertices, no two of which are adjacent. a subset of vertices, no two of which are adjacent.
This is one of Karp's 21 NP-complete problems. This is one of Karp's 21 NP-complete problems.
""" """
def __init__(self, graph, weights): def __init__(self, graph, weights):
self.graph = graph self.graph = graph
self.weights = weights self.weights = weights
@@ -102,13 +108,14 @@ class MaxWeightStableSetInstance(Instance):
nodes = list(self.graph.nodes) nodes = list(self.graph.nodes)
model = pe.ConcreteModel() model = pe.ConcreteModel()
model.x = pe.Var(nodes, domain=pe.Binary) model.x = pe.Var(nodes, domain=pe.Binary)
model.OBJ = pe.Objective(expr=sum(model.x[v] * self.weights[v] for v in nodes), model.OBJ = pe.Objective(
sense=pe.maximize) expr=sum(model.x[v] * self.weights[v] for v in nodes), sense=pe.maximize
)
model.clique_eqs = pe.ConstraintList() model.clique_eqs = pe.ConstraintList()
for clique in nx.find_cliques(self.graph): for clique in nx.find_cliques(self.graph):
model.clique_eqs.add(sum(model.x[i] for i in clique) <= 1) model.clique_eqs.add(sum(model.x[i] for i in clique) <= 1)
return model return model
def get_instance_features(self): def get_instance_features(self):
return np.ones(0) return np.ones(0)

1
miplearn/problems/tests/__init__.py
View File

@@ -1,4 +1,3 @@
# MIPLearn: Extensible Framework for Learning-Enhanced Mixed-Integer Optimization # MIPLearn: Extensible Framework for Learning-Enhanced Mixed-Integer Optimization
# Copyright (C) 2020, UChicago Argonne, LLC. All rights reserved. # Copyright (C) 2020, UChicago Argonne, LLC. All rights reserved.
# Released under the modified BSD license. See COPYING.md for more details. # Released under the modified BSD license. See COPYING.md for more details.

19
miplearn/problems/tests/test_knapsack.py
View File

@@ -9,17 +9,18 @@ import numpy as np
def test_knapsack_generator(): def test_knapsack_generator():
gen = MultiKnapsackGenerator(n=randint(low=100, high=101), gen = MultiKnapsackGenerator(
m=randint(low=30, high=31), n=randint(low=100, high=101),
w=randint(low=0, high=1000), m=randint(low=30, high=31),
K=randint(low=500, high=501), w=randint(low=0, high=1000),
u=uniform(loc=1.0, scale=1.0), K=randint(low=500, high=501),
alpha=uniform(loc=0.50, scale=0.0), u=uniform(loc=1.0, scale=1.0),
) alpha=uniform(loc=0.50, scale=0.0),
)
instances = gen.generate(100) instances = gen.generate(100)
w_sum = sum(instance.weights for instance in instances) / len(instances) w_sum = sum(instance.weights for instance in instances) / len(instances)
p_sum = sum(instance.prices for instance in instances) / len(instances) p_sum = sum(instance.prices for instance in instances) / len(instances)
b_sum = sum(instance.capacities for instance in instances) / len(instances) b_sum = sum(instance.capacities for instance in instances) / len(instances)
assert round(np.mean(w_sum), -1) == 500. assert round(np.mean(w_sum), -1) == 500.0
# assert round(np.mean(p_sum), -1) == 1200. # flaky # assert round(np.mean(p_sum), -1) == 1200. # flaky
assert round(np.mean(b_sum), -3) == 25000. assert round(np.mean(b_sum), -3) == 25000.0

38
miplearn/problems/tests/test_stab.py
View File

@@ -11,36 +11,42 @@ from scipy.stats import uniform, randint
def test_stab(): def test_stab():
graph = nx.cycle_graph(5) graph = nx.cycle_graph(5)
weights = [1., 1., 1., 1., 1.] weights = [1.0, 1.0, 1.0, 1.0, 1.0]
instance = MaxWeightStableSetInstance(graph, weights) instance = MaxWeightStableSetInstance(graph, weights)
solver = LearningSolver() solver = LearningSolver()
solver.solve(instance) solver.solve(instance)
assert instance.lower_bound == 2. assert instance.lower_bound == 2.0
def test_stab_generator_fixed_graph(): def test_stab_generator_fixed_graph():
np.random.seed(42) np.random.seed(42)
from miplearn.problems.stab import MaxWeightStableSetGenerator from miplearn.problems.stab import MaxWeightStableSetGenerator
gen = MaxWeightStableSetGenerator(w=uniform(loc=50., scale=10.),
n=randint(low=10, high=11), gen = MaxWeightStableSetGenerator(
p=uniform(loc=0.05, scale=0.), w=uniform(loc=50.0, scale=10.0),
fix_graph=True) n=randint(low=10, high=11),
p=uniform(loc=0.05, scale=0.0),
fix_graph=True,
)
instances = gen.generate(1_000) instances = gen.generate(1_000)
weights = np.array([instance.weights for instance in instances]) weights = np.array([instance.weights for instance in instances])
weights_avg_actual = np.round(np.average(weights, axis=0)) weights_avg_actual = np.round(np.average(weights, axis=0))
weights_avg_expected = [55.0] * 10 weights_avg_expected = [55.0] * 10
assert list(weights_avg_actual) == weights_avg_expected assert list(weights_avg_actual) == weights_avg_expected
def test_stab_generator_random_graph(): def test_stab_generator_random_graph():
np.random.seed(42) np.random.seed(42)
from miplearn.problems.stab import MaxWeightStableSetGenerator from miplearn.problems.stab import MaxWeightStableSetGenerator
gen = MaxWeightStableSetGenerator(w=uniform(loc=50., scale=10.),
n=randint(low=30, high=41), gen = MaxWeightStableSetGenerator(
p=uniform(loc=0.5, scale=0.), w=uniform(loc=50.0, scale=10.0),
fix_graph=False) n=randint(low=30, high=41),
p=uniform(loc=0.5, scale=0.0),
fix_graph=False,
)
instances = gen.generate(1_000) instances = gen.generate(1_000)
n_nodes = [instance.graph.number_of_nodes() for instance in instances] n_nodes = [instance.graph.number_of_nodes() for instance in instances]
n_edges = [instance.graph.number_of_edges() for instance in instances] n_edges = [instance.graph.number_of_edges() for instance in instances]
assert np.round(np.mean(n_nodes)) == 35. assert np.round(np.mean(n_nodes)) == 35.0
assert np.round(np.mean(n_edges), -1) == 300. assert np.round(np.mean(n_edges), -1) == 300.0

48
miplearn/problems/tests/test_tsp.py
View File

@@ -11,11 +11,13 @@ from scipy.stats import uniform, randint
def test_generator(): def test_generator():
instances = TravelingSalesmanGenerator(x=uniform(loc=0.0, scale=1000.0), instances = TravelingSalesmanGenerator(
y=uniform(loc=0.0, scale=1000.0), x=uniform(loc=0.0, scale=1000.0),
n=randint(low=100, high=101), y=uniform(loc=0.0, scale=1000.0),
gamma=uniform(loc=0.95, scale=0.1), n=randint(low=100, high=101),
fix_cities=True).generate(100) gamma=uniform(loc=0.95, scale=0.1),
fix_cities=True,
).generate(100)
assert len(instances) == 100 assert len(instances) == 100
assert instances[0].n_cities == 100 assert instances[0].n_cities == 100
assert norm(instances[0].distances - instances[0].distances.T) < 1e-6 assert norm(instances[0].distances - instances[0].distances.T) < 1e-6
@@ -25,14 +27,16 @@ def test_generator():
def test_instance(): def test_instance():
n_cities = 4 n_cities = 4
distances = np.array([ distances = np.array(
[0., 1., 2., 1.], [
[1., 0., 1., 2.], [0.0, 1.0, 2.0, 1.0],
[2., 1., 0., 1.], [1.0, 0.0, 1.0, 2.0],
[1., 2., 1., 0.], [2.0, 1.0, 0.0, 1.0],
]) [1.0, 2.0, 1.0, 0.0],
]
)
instance = TravelingSalesmanInstance(n_cities, distances) instance = TravelingSalesmanInstance(n_cities, distances)
for solver_name in ['gurobi']: for solver_name in ["gurobi"]:
solver = LearningSolver(solver=solver_name) solver = LearningSolver(solver=solver_name)
solver.solve(instance) solver.solve(instance)
x = instance.solution["x"] x = instance.solution["x"]
@@ -48,17 +52,19 @@ def test_instance():
def test_subtour(): def test_subtour():
n_cities = 6 n_cities = 6
cities = np.array([ cities = np.array(
[0., 0.], [
[1., 0.], [0.0, 0.0],
[2., 0.], [1.0, 0.0],
[3., 0.], [2.0, 0.0],
[0., 1.], [3.0, 0.0],
[3., 1.], [0.0, 1.0],
]) [3.0, 1.0],
]
)
distances = squareform(pdist(cities)) distances = squareform(pdist(cities))
instance = TravelingSalesmanInstance(n_cities, distances) instance = TravelingSalesmanInstance(n_cities, distances)
for solver_name in ['gurobi']: for solver_name in ["gurobi"]:
solver = LearningSolver(solver=solver_name) solver = LearningSolver(solver=solver_name)
solver.solve(instance) solver.solve(instance)
assert hasattr(instance, "found_violated_lazy_constraints") assert hasattr(instance, "found_violated_lazy_constraints")

121
miplearn/problems/tsp.py
View File

@@ -13,41 +13,44 @@ import random
class ChallengeA: class ChallengeA:
def __init__(self, def __init__(
seed=42, self,
n_training_instances=500, seed=42,
n_test_instances=50, n_training_instances=500,
): n_test_instances=50,
):
np.random.seed(seed) np.random.seed(seed)
self.generator = TravelingSalesmanGenerator(x=uniform(loc=0.0, scale=1000.0), self.generator = TravelingSalesmanGenerator(
y=uniform(loc=0.0, scale=1000.0), x=uniform(loc=0.0, scale=1000.0),
n=randint(low=350, high=351), y=uniform(loc=0.0, scale=1000.0),
gamma=uniform(loc=0.95, scale=0.1), n=randint(low=350, high=351),
fix_cities=True, gamma=uniform(loc=0.95, scale=0.1),
round=True, fix_cities=True,
) round=True,
)
np.random.seed(seed + 1) np.random.seed(seed + 1)
self.training_instances = self.generator.generate(n_training_instances) self.training_instances = self.generator.generate(n_training_instances)
np.random.seed(seed + 2) np.random.seed(seed + 2)
self.test_instances = self.generator.generate(n_test_instances) self.test_instances = self.generator.generate(n_test_instances)
class TravelingSalesmanGenerator: class TravelingSalesmanGenerator:
"""Random generator for the Traveling Salesman Problem.""" """Random generator for the Traveling Salesman Problem."""
def __init__(self, def __init__(
x=uniform(loc=0.0, scale=1000.0), self,
y=uniform(loc=0.0, scale=1000.0), x=uniform(loc=0.0, scale=1000.0),
n=randint(low=100, high=101), y=uniform(loc=0.0, scale=1000.0),
gamma=uniform(loc=1.0, scale=0.0), n=randint(low=100, high=101),
fix_cities=True, gamma=uniform(loc=1.0, scale=0.0),
round=True, fix_cities=True,
): round=True,
):
"""Initializes the problem generator. """Initializes the problem generator.
Initially, the generator creates n cities (x_1,y_1),...,(x_n,y_n) where n, x_i and y_i are Initially, the generator creates n cities (x_1,y_1),...,(x_n,y_n) where n, x_i and y_i are
sampled independently from the provided probability distributions `n`, `x` and `y`. For each sampled independently from the provided probability distributions `n`, `x` and `y`. For each
(unordered) pair of cities (i,j), the distance d[i,j] between them is set to: (unordered) pair of cities (i,j), the distance d[i,j] between them is set to:
@@ -58,7 +61,7 @@ class TravelingSalesmanGenerator:
If fix_cities=True, the list of cities is kept the same for all generated instances. The If fix_cities=True, the list of cities is kept the same for all generated instances. The
gamma values, and therefore also the distances, are still different. gamma values, and therefore also the distances, are still different.
By default, all distances d[i,j] are rounded to the nearest integer. If `round=False` By default, all distances d[i,j] are rounded to the nearest integer. If `round=False`
is provided, this rounding will be disabled. is provided, this rounding will be disabled.
@@ -79,19 +82,22 @@ class TravelingSalesmanGenerator:
assert isinstance(x, rv_frozen), "x should be a SciPy probability distribution" assert isinstance(x, rv_frozen), "x should be a SciPy probability distribution"
assert isinstance(y, rv_frozen), "y should be a SciPy probability distribution" assert isinstance(y, rv_frozen), "y should be a SciPy probability distribution"
assert isinstance(n, rv_frozen), "n should be a SciPy probability distribution" assert isinstance(n, rv_frozen), "n should be a SciPy probability distribution"
assert isinstance(gamma, rv_frozen), "gamma should be a SciPy probability distribution" assert isinstance(
gamma,
rv_frozen,
), "gamma should be a SciPy probability distribution"
self.x = x self.x = x
self.y = y self.y = y
self.n = n self.n = n
self.gamma = gamma self.gamma = gamma
self.round = round self.round = round
if fix_cities: if fix_cities:
self.fixed_n, self.fixed_cities = self._generate_cities() self.fixed_n, self.fixed_cities = self._generate_cities()
else: else:
self.fixed_n = None self.fixed_n = None
self.fixed_cities = None self.fixed_cities = None
def generate(self, n_samples): def generate(self, n_samples):
def _sample(): def _sample():
if self.fixed_cities is not None: if self.fixed_cities is not None:
@@ -103,54 +109,62 @@ class TravelingSalesmanGenerator:
if self.round: if self.round:
distances = distances.round() distances = distances.round()
return TravelingSalesmanInstance(n, distances) return TravelingSalesmanInstance(n, distances)
return [_sample() for _ in range(n_samples)] return [_sample() for _ in range(n_samples)]
def _generate_cities(self): def _generate_cities(self):
n = self.n.rvs() n = self.n.rvs()
cities = np.array([(self.x.rvs(), self.y.rvs()) for _ in range(n)]) cities = np.array([(self.x.rvs(), self.y.rvs()) for _ in range(n)])
return n, cities return n, cities
class TravelingSalesmanInstance(Instance): class TravelingSalesmanInstance(Instance):
"""An instance ot the Traveling Salesman Problem. """An instance ot the Traveling Salesman Problem.
Given a list of cities and the distance between each pair of cities, the problem asks for the Given a list of cities and the distance between each pair of cities, the problem asks for the
shortest route starting at the first city, visiting each other city exactly once, then shortest route starting at the first city, visiting each other city exactly once, then
returning to the first city. This problem is a generalization of the Hamiltonian path problem, returning to the first city. This problem is a generalization of the Hamiltonian path problem,
one of Karp's 21 NP-complete problems. one of Karp's 21 NP-complete problems.
""" """
def __init__(self, n_cities, distances): def __init__(self, n_cities, distances):
assert isinstance(distances, np.ndarray) assert isinstance(distances, np.ndarray)
assert distances.shape == (n_cities, n_cities) assert distances.shape == (n_cities, n_cities)
self.n_cities = n_cities self.n_cities = n_cities
self.distances = distances self.distances = distances
def to_model(self): def to_model(self):
model = pe.ConcreteModel() model = pe.ConcreteModel()
model.edges = edges = [(i,j) model.edges = edges = [
for i in range(self.n_cities) (i, j) for i in range(self.n_cities) for j in range(i + 1, self.n_cities)
for j in range(i+1, self.n_cities)] ]
model.x = pe.Var(edges, domain=pe.Binary) model.x = pe.Var(edges, domain=pe.Binary)
model.obj = pe.Objective(expr=sum(model.x[i,j] * self.distances[i,j] model.obj = pe.Objective(
for (i,j) in edges), expr=sum(model.x[i, j] * self.distances[i, j] for (i, j) in edges),
sense=pe.minimize) sense=pe.minimize,
)
model.eq_degree = pe.ConstraintList() model.eq_degree = pe.ConstraintList()
model.eq_subtour = pe.ConstraintList() model.eq_subtour = pe.ConstraintList()
for i in range(self.n_cities): for i in range(self.n_cities):
model.eq_degree.add(sum(model.x[min(i,j), max(i,j)] model.eq_degree.add(
for j in range(self.n_cities) if i != j) == 2) sum(
model.x[min(i, j), max(i, j)]
for j in range(self.n_cities)
if i != j
)
== 2
)
return model return model
def get_instance_features(self): def get_instance_features(self):
return np.array([1]) return np.array([1])
def get_variable_features(self, var_name, index): def get_variable_features(self, var_name, index):
return np.array([1]) return np.array([1])
def get_variable_category(self, var_name, index): def get_variable_category(self, var_name, index):
return index return index
def find_violated_lazy_constraints(self, model): def find_violated_lazy_constraints(self, model):
selected_edges = [e for e in model.edges if model.x[e].value > 0.5] selected_edges = [e for e in model.edges if model.x[e].value > 0.5]
graph = nx.Graph() graph = nx.Graph()
@@ -161,15 +175,18 @@ class TravelingSalesmanInstance(Instance):
if len(c) < self.n_cities: if len(c) < self.n_cities:
violations += [c] violations += [c]
return violations return violations
def build_lazy_constraint(self, model, component): def build_lazy_constraint(self, model, component):
cut_edges = [e for e in model.edges cut_edges = [
if (e[0] in component and e[1] not in component) or e
(e[0] not in component and e[1] in component)] for e in model.edges
if (e[0] in component and e[1] not in component)
or (e[0] not in component and e[1] in component)
]
return model.eq_subtour.add(sum(model.x[e] for e in cut_edges) >= 2) return model.eq_subtour.add(sum(model.x[e] for e in cut_edges) >= 2)
def find_violated_user_cuts(self, model): def find_violated_user_cuts(self, model):
return self.find_violated_lazy_constraints(model) return self.find_violated_lazy_constraints(model)
def build_user_cut(self, model, violation): def build_user_cut(self, model, violation):
return self.build_lazy_constraint(model, violation) return self.build_lazy_constraint(model, violation)

52
miplearn/solvers/gurobi.py
View File

@@ -13,10 +13,11 @@ logger = logging.getLogger(__name__)
class GurobiSolver(InternalSolver): class GurobiSolver(InternalSolver):
def __init__(self, def __init__(
params=None, self,
lazy_cb_frequency=1, params=None,
): lazy_cb_frequency=1,
):
""" """
An InternalSolver backed by Gurobi's Python API (without Pyomo). An InternalSolver backed by Gurobi's Python API (without Pyomo).
@@ -33,6 +34,7 @@ class GurobiSolver(InternalSolver):
if params is None: if params is None:
params = {} params = {}
from gurobipy import GRB from gurobipy import GRB
self.GRB = GRB self.GRB = GRB
self.instance = None self.instance = None
self.model = None self.model = None
@@ -44,8 +46,7 @@ class GurobiSolver(InternalSolver):
if lazy_cb_frequency == 1: if lazy_cb_frequency == 1:
self.lazy_cb_where = [self.GRB.Callback.MIPSOL] self.lazy_cb_where = [self.GRB.Callback.MIPSOL]
else: else:
self.lazy_cb_where = [self.GRB.Callback.MIPSOL, self.lazy_cb_where = [self.GRB.Callback.MIPSOL, self.GRB.Callback.MIPNODE]
self.GRB.Callback.MIPNODE]
def set_instance(self, instance, model=None): def set_instance(self, instance, model=None):
self._raise_if_callback() self._raise_if_callback()
@@ -70,14 +71,15 @@ class GurobiSolver(InternalSolver):
idx = [0] idx = [0]
else: else:
name = m.group(1) name = m.group(1)
idx = tuple(int(k) if k.isdecimal() else k idx = tuple(
for k in m.group(2).split(",")) int(k) if k.isdecimal() else k for k in m.group(2).split(",")
)
if len(idx) == 1: if len(idx) == 1:
idx = idx[0] idx = idx[0]
if name not in self._all_vars: if name not in self._all_vars:
self._all_vars[name] = {} self._all_vars[name] = {}
self._all_vars[name][idx] = var self._all_vars[name][idx] = var
if var.vtype != 'C': if var.vtype != "C":
if name not in self._bin_vars: if name not in self._bin_vars:
self._bin_vars[name] = {} self._bin_vars[name] = {}
self._bin_vars[name][idx] = var self._bin_vars[name][idx] = var
@@ -103,15 +105,9 @@ class GurobiSolver(InternalSolver):
for (idx, var) in vardict.items(): for (idx, var) in vardict.items():
var.vtype = self.GRB.BINARY var.vtype = self.GRB.BINARY
log = streams[0].getvalue() log = streams[0].getvalue()
return { return {"Optimal value": self.model.objVal, "Log": log}
"Optimal value": self.model.objVal,
"Log": log
}
def solve(self, def solve(self, tee=False, iteration_cb=None, lazy_cb=None):
tee=False,
iteration_cb=None,
lazy_cb=None):
self._raise_if_callback() self._raise_if_callback()
def cb_wrapper(cb_model, cb_where): def cb_wrapper(cb_model, cb_where):
@@ -133,7 +129,7 @@ class GurobiSolver(InternalSolver):
if tee: if tee:
streams += [sys.stdout] streams += [sys.stdout]
if iteration_cb is None: if iteration_cb is None:
iteration_cb = lambda : False iteration_cb = lambda: False
while True: while True:
logger.debug("Solving MIP...") logger.debug("Solving MIP...")
with RedirectOutput(streams): with RedirectOutput(streams):
@@ -187,7 +183,9 @@ class GurobiSolver(InternalSolver):
elif self.cb_where is None: elif self.cb_where is None:
return var.x return var.x
else: else:
raise Exception("get_value cannot be called from cb_where=%s" % self.cb_where) raise Exception(
"get_value cannot be called from cb_where=%s" % self.cb_where
)
def get_variables(self): def get_variables(self):
self._raise_if_callback() self._raise_if_callback()
@@ -220,8 +218,10 @@ class GurobiSolver(InternalSolver):
if value is not None: if value is not None:
count_fixed += 1 count_fixed += 1
self._all_vars[varname][idx].start = value self._all_vars[varname][idx].start = value
logger.info("Setting start values for %d variables (out of %d)" % logger.info(
(count_fixed, count_total)) "Setting start values for %d variables (out of %d)"
% (count_fixed, count_total)
)
def clear_warm_start(self): def clear_warm_start(self):
self._raise_if_callback() self._raise_if_callback()
@@ -248,10 +248,7 @@ class GurobiSolver(InternalSolver):
def extract_constraint(self, cid): def extract_constraint(self, cid):
self._raise_if_callback() self._raise_if_callback()
constr = self.model.getConstrByName(cid) constr = self.model.getConstrByName(cid)
cobj = (self.model.getRow(constr), cobj = (self.model.getRow(constr), constr.sense, constr.RHS, constr.ConstrName)
constr.sense,
constr.RHS,
constr.ConstrName)
self.model.remove(constr) self.model.remove(constr)
return cobj return cobj
@@ -316,7 +313,7 @@ class GurobiSolver(InternalSolver):
value = matches[0] value = matches[0]
return value return value
def __getstate__(self): def __getstate__(self):
return { return {
"params": self.params, "params": self.params,
"lazy_cb_where": self.lazy_cb_where, "lazy_cb_where": self.lazy_cb_where,
@@ -324,6 +321,7 @@ class GurobiSolver(InternalSolver):
def __setstate__(self, state): def __setstate__(self, state):
from gurobipy import GRB from gurobipy import GRB
self.params = state["params"] self.params = state["params"]
self.lazy_cb_where = state["lazy_cb_where"] self.lazy_cb_where = state["lazy_cb_where"]
self.GRB = GRB self.GRB = GRB
@@ -331,4 +329,4 @@ class GurobiSolver(InternalSolver):
self.model = None self.model = None
self._all_vars = None self._all_vars = None
self._bin_vars = None self._bin_vars = None
self.cb_where = None self.cb_where = None

1
miplearn/solvers/internal.py
View File

@@ -222,4 +222,3 @@ class InternalSolver(ABC):
for idx in indices: for idx in indices:
solution[var][idx] = 0.0 solution[var][idx] = 0.0
return solution return solution

94
miplearn/solvers/learning.py
View File

@@ -12,10 +12,12 @@ from copy import deepcopy
from typing import Optional, List from typing import Optional, List
from p_tqdm import p_map from p_tqdm import p_map
from .. import (ObjectiveValueComponent, from .. import (
PrimalSolutionComponent, ObjectiveValueComponent,
DynamicLazyConstraintsComponent, PrimalSolutionComponent,
UserCutsComponent) DynamicLazyConstraintsComponent,
UserCutsComponent,
)
from .pyomo.cplex import CplexPyomoSolver from .pyomo.cplex import CplexPyomoSolver
from .pyomo.gurobi import GurobiPyomoSolver from .pyomo.gurobi import GurobiPyomoSolver
@@ -43,16 +45,18 @@ def _parallel_solve(idx):
class LearningSolver: class LearningSolver:
def __init__(self, def __init__(
components=None, self,
gap_tolerance=1e-4, components=None,
mode="exact", gap_tolerance=1e-4,
solver="gurobi", mode="exact",
threads=None, solver="gurobi",
time_limit=None, threads=None,
node_limit=None, time_limit=None,
solve_lp_first=True, node_limit=None,
use_lazy_cb=False): solve_lp_first=True,
use_lazy_cb=False,
):
""" """
Mixed-Integer Linear Programming (MIP) solver that extracts information Mixed-Integer Linear Programming (MIP) solver that extracts information
from previous runs and uses Machine Learning methods to accelerate the from previous runs and uses Machine Learning methods to accelerate the
@@ -142,28 +146,30 @@ class LearningSolver:
solver.set_node_limit(self.node_limit) solver.set_node_limit(self.node_limit)
return solver return solver
def solve(self, def solve(
instance, self,
model=None, instance,
output="", model=None,
tee=False): output="",
tee=False,
):
""" """
Solves the given instance. If trained machine-learning models are Solves the given instance. If trained machine-learning models are
available, they will be used to accelerate the solution process. available, they will be used to accelerate the solution process.
The argument `instance` may be either an Instance object or a The argument `instance` may be either an Instance object or a
filename pointing to a pickled Instance object. filename pointing to a pickled Instance object.
This method modifies the instance object. Specifically, the following This method modifies the instance object. Specifically, the following
properties are set: properties are set:
- instance.lp_solution - instance.lp_solution
- instance.lp_value - instance.lp_value
- instance.lower_bound - instance.lower_bound
- instance.upper_bound - instance.upper_bound
- instance.solution - instance.solution
- instance.solver_log - instance.solver_log
Additional solver components may set additional properties. Please Additional solver components may set additional properties. Please
see their documentation for more details. If a filename is provided, see their documentation for more details. If a filename is provided,
then the file is modified in-place. That is, the original file is then the file is modified in-place. That is, the original file is
@@ -197,7 +203,7 @@ class LearningSolver:
"Predicted UB". See the documentation of each component for more "Predicted UB". See the documentation of each component for more
details. details.
""" """
filename = None filename = None
fileformat = None fileformat = None
if isinstance(instance, str): if isinstance(instance, str):
@@ -211,7 +217,7 @@ class LearningSolver:
fileformat = "pickle" fileformat = "pickle"
with open(filename, "rb") as file: with open(filename, "rb") as file:
instance = pickle.load(file) instance = pickle.load(file)
if model is None: if model is None:
model = instance.to_model() model = instance.to_model()
@@ -248,9 +254,11 @@ class LearningSolver:
lazy_cb = lazy_cb_wrapper lazy_cb = lazy_cb_wrapper
logger.info("Solving MILP...") logger.info("Solving MILP...")
results = self.internal_solver.solve(tee=tee, results = self.internal_solver.solve(
iteration_cb=iteration_cb, tee=tee,
lazy_cb=lazy_cb) iteration_cb=iteration_cb,
lazy_cb=lazy_cb,
)
results["LP value"] = instance.lp_value results["LP value"] = instance.lp_value
# Read MIP solution and bounds # Read MIP solution and bounds
@@ -262,7 +270,7 @@ class LearningSolver:
logger.debug("Calling after_solve callbacks...") logger.debug("Calling after_solve callbacks...")
for component in self.components.values(): for component in self.components.values():
component.after_solve(self, instance, model, results) component.after_solve(self, instance, model, results)
if filename is not None and output is not None: if filename is not None and output is not None:
output_filename = output output_filename = output
if len(output) == 0: if len(output) == 0:
@@ -280,36 +288,38 @@ class LearningSolver:
def parallel_solve(self, instances, n_jobs=4, label="Solve", output=[]): def parallel_solve(self, instances, n_jobs=4, label="Solve", output=[]):
""" """
Solves multiple instances in parallel. Solves multiple instances in parallel.
This method is equivalent to calling `solve` for each item on the list, This method is equivalent to calling `solve` for each item on the list,
but it processes multiple instances at the same time. Like `solve`, this but it processes multiple instances at the same time. Like `solve`, this
method modifies each instance in place. Also like `solve`, a list of method modifies each instance in place. Also like `solve`, a list of
filenames may be provided. filenames may be provided.
Parameters Parameters
---------- ----------
instances: [miplearn.Instance] or [str] instances: [miplearn.Instance] or [str]
The instances to be solved The instances to be solved
n_jobs: int n_jobs: int
Number of instances to solve in parallel at a time. Number of instances to solve in parallel at a time.
Returns Returns
------- -------
Returns a list of dictionaries, with one entry for each provided instance. Returns a list of dictionaries, with one entry for each provided instance.
This dictionary is the same you would obtain by calling: This dictionary is the same you would obtain by calling:
[solver.solve(p) for p in instances] [solver.solve(p) for p in instances]
""" """
self.internal_solver = None self.internal_solver = None
self._silence_miplearn_logger() self._silence_miplearn_logger()
SOLVER[0] = self SOLVER[0] = self
OUTPUTS[0] = output OUTPUTS[0] = output
INSTANCES[0] = instances INSTANCES[0] = instances
results = p_map(_parallel_solve, results = p_map(
list(range(len(instances))), _parallel_solve,
num_cpus=n_jobs, list(range(len(instances))),
desc=label) num_cpus=n_jobs,
desc=label,
)
stats = [] stats = []
for (idx, (s, instance)) in enumerate(results): for (idx, (s, instance)) in enumerate(results):
stats.append(s) stats.append(s)
@@ -330,12 +340,12 @@ class LearningSolver:
def _silence_miplearn_logger(self): def _silence_miplearn_logger(self):
miplearn_logger = logging.getLogger("miplearn") miplearn_logger = logging.getLogger("miplearn")
self.prev_log_level = miplearn_logger.getEffectiveLevel() self.prev_log_level = miplearn_logger.getEffectiveLevel()
miplearn_logger.setLevel(logging.WARNING) miplearn_logger.setLevel(logging.WARNING)
def _restore_miplearn_logger(self): def _restore_miplearn_logger(self):
miplearn_logger = logging.getLogger("miplearn") miplearn_logger = logging.getLogger("miplearn")
miplearn_logger.setLevel(self.prev_log_level) miplearn_logger.setLevel(self.prev_log_level)
def __getstate__(self): def __getstate__(self):
self.internal_solver = None self.internal_solver = None
return self.__dict__ return self.__dict__

32
miplearn/solvers/pyomo/base.py
View File

@@ -81,8 +81,10 @@ class BasePyomoSolver(InternalSolver):
count_fixed += 1 count_fixed += 1
if count_fixed > 0: if count_fixed > 0:
self._is_warm_start_available = True self._is_warm_start_available = True
logger.info("Setting start values for %d variables (out of %d)" % logger.info(
(count_fixed, count_total)) "Setting start values for %d variables (out of %d)"
% (count_fixed, count_total)
)
def clear_warm_start(self): def clear_warm_start(self):
for var in self._all_vars: for var in self._all_vars:
@@ -134,17 +136,19 @@ class BasePyomoSolver(InternalSolver):
count_fixed += 1 count_fixed += 1
var[index].fix(solution[varname][index]) var[index].fix(solution[varname][index])
self._pyomo_solver.update_var(var[index]) self._pyomo_solver.update_var(var[index])
logger.info("Fixing values for %d variables (out of %d)" % logger.info(
(count_fixed, count_total)) "Fixing values for %d variables (out of %d)"
% (
count_fixed,
count_total,
)
)
def add_constraint(self, constraint): def add_constraint(self, constraint):
self._pyomo_solver.add_constraint(constraint) self._pyomo_solver.add_constraint(constraint)
self._update_constrs() self._update_constrs()
def solve(self, def solve(self, tee=False, iteration_cb=None, lazy_cb=None):
tee=False,
iteration_cb=None,
lazy_cb=None):
if lazy_cb is not None: if lazy_cb is not None:
raise Exception("lazy callback not supported") raise Exception("lazy callback not supported")
total_wallclock_time = 0 total_wallclock_time = 0
@@ -158,8 +162,10 @@ class BasePyomoSolver(InternalSolver):
while True: while True:
logger.debug("Solving MIP...") logger.debug("Solving MIP...")
with RedirectOutput(streams): with RedirectOutput(streams):
results = self._pyomo_solver.solve(tee=True, results = self._pyomo_solver.solve(
warmstart=self._is_warm_start_available) tee=True,
warmstart=self._is_warm_start_available,
)
total_wallclock_time += results["Solver"][0]["Wallclock time"] total_wallclock_time += results["Solver"][0]["Wallclock time"]
should_repeat = iteration_cb() should_repeat = iteration_cb()
if not should_repeat: if not should_repeat:
@@ -192,9 +198,7 @@ class BasePyomoSolver(InternalSolver):
return value return value
def _extract_node_count(self, log): def _extract_node_count(self, log):
return int(self.__extract(log, return int(self.__extract(log, self._get_node_count_regexp(), default=1))
self._get_node_count_regexp(),
default=1))
def set_threads(self, threads): def set_threads(self, threads):
key = self._get_threads_option_name() key = self._get_threads_option_name()
@@ -249,4 +253,4 @@ class BasePyomoSolver(InternalSolver):
raise Exception("not implemented") raise Exception("not implemented")
def get_constraint_slacks(self): def get_constraint_slacks(self):
raise Exception("not implemented") raise Exception("not implemented")

2
miplearn/solvers/pyomo/cplex.py
View File

@@ -20,7 +20,7 @@ class CplexPyomoSolver(BasePyomoSolver):
{"mip_display": 5} to increase the log verbosity. {"mip_display": 5} to increase the log verbosity.
""" """
super().__init__() super().__init__()
self._pyomo_solver = pe.SolverFactory('cplex_persistent') self._pyomo_solver = pe.SolverFactory("cplex_persistent")
self._pyomo_solver.options["randomseed"] = randint(low=0, high=1000).rvs() self._pyomo_solver.options["randomseed"] = randint(low=0, high=1000).rvs()
self._pyomo_solver.options["mip_display"] = 4 self._pyomo_solver.options["mip_display"] = 4
if options is not None: if options is not None:

6
miplearn/solvers/pyomo/gurobi.py
View File

@@ -15,8 +15,7 @@ logger = logging.getLogger(__name__)
class GurobiPyomoSolver(BasePyomoSolver): class GurobiPyomoSolver(BasePyomoSolver):
def __init__(self, def __init__(self, options=None):
options=None):
""" """
Creates a new Gurobi solver, accessed through Pyomo. Creates a new Gurobi solver, accessed through Pyomo.
@@ -27,7 +26,7 @@ class GurobiPyomoSolver(BasePyomoSolver):
{"Threads": 4} to set the number of threads. {"Threads": 4} to set the number of threads.
""" """
super().__init__() super().__init__()
self._pyomo_solver = pe.SolverFactory('gurobi_persistent') self._pyomo_solver = pe.SolverFactory("gurobi_persistent")
self._pyomo_solver.options["Seed"] = randint(low=0, high=1000).rvs() self._pyomo_solver.options["Seed"] = randint(low=0, high=1000).rvs()
if options is not None: if options is not None:
for (key, value) in options.items(): for (key, value) in options.items():
@@ -56,6 +55,7 @@ class GurobiPyomoSolver(BasePyomoSolver):
def set_branching_priorities(self, priorities): def set_branching_priorities(self, priorities):
from gurobipy import GRB from gurobipy import GRB
for varname in priorities.keys(): for varname in priorities.keys():
var = self._varname_to_var[varname] var = self._varname_to_var[varname]
for (index, priority) in priorities[varname].items(): for (index, priority) in priorities[varname].items():

16
miplearn/solvers/tests/__init__.py
View File

@@ -9,20 +9,22 @@ from miplearn.problems.knapsack import KnapsackInstance, GurobiKnapsackInstance
def _get_instance(solver): def _get_instance(solver):
def _is_subclass_or_instance(solver, parentClass): def _is_subclass_or_instance(solver, parentClass):
return isinstance(solver, parentClass) or (isclass(solver) and issubclass(solver, parentClass)) return isinstance(solver, parentClass) or (
isclass(solver) and issubclass(solver, parentClass)
)
if _is_subclass_or_instance(solver, BasePyomoSolver): if _is_subclass_or_instance(solver, BasePyomoSolver):
return KnapsackInstance( return KnapsackInstance(
weights=[23., 26., 20., 18.], weights=[23.0, 26.0, 20.0, 18.0],
prices=[505., 352., 458., 220.], prices=[505.0, 352.0, 458.0, 220.0],
capacity=67., capacity=67.0,
) )
if _is_subclass_or_instance(solver, GurobiSolver): if _is_subclass_or_instance(solver, GurobiSolver):
return GurobiKnapsackInstance( return GurobiKnapsackInstance(
weights=[23., 26., 20., 18.], weights=[23.0, 26.0, 20.0, 18.0],
prices=[505., 352., 458., 220.], prices=[505.0, 352.0, 458.0, 220.0],
capacity=67., capacity=67.0,
) )
assert False assert False

49
miplearn/solvers/tests/test_internal_solver.py
View File

@@ -16,6 +16,7 @@ logger = logging.getLogger(__name__)
def test_redirect_output(): def test_redirect_output():
import sys import sys
original_stdout = sys.stdout original_stdout = sys.stdout
io = StringIO() io = StringIO()
with RedirectOutput([io]): with RedirectOutput([io]):
@@ -31,36 +32,42 @@ def test_internal_solver_warm_starts():
model = instance.to_model() model = instance.to_model()
solver = solver_class() solver = solver_class()
solver.set_instance(instance, model) solver.set_instance(instance, model)
solver.set_warm_start({ solver.set_warm_start(
"x": { {
0: 1.0, "x": {
1: 0.0, 0: 1.0,
2: 0.0, 1: 0.0,
3: 1.0, 2: 0.0,
3: 1.0,
}
} }
}) )
stats = solver.solve(tee=True) stats = solver.solve(tee=True)
assert stats["Warm start value"] == 725.0 assert stats["Warm start value"] == 725.0
solver.set_warm_start({ solver.set_warm_start(
"x": { {
0: 1.0, "x": {
1: 1.0, 0: 1.0,
2: 1.0, 1: 1.0,
3: 1.0, 2: 1.0,
3: 1.0,
}
} }
}) )
stats = solver.solve(tee=True) stats = solver.solve(tee=True)
assert stats["Warm start value"] is None assert stats["Warm start value"] is None
solver.fix({ solver.fix(
"x": { {
0: 1.0, "x": {
1: 0.0, 0: 1.0,
2: 0.0, 1: 0.0,
3: 1.0, 2: 0.0,
3: 1.0,
}
} }
}) )
stats = solver.solve(tee=True) stats = solver.solve(tee=True)
assert stats["Lower bound"] == 725.0 assert stats["Lower bound"] == 725.0
assert stats["Upper bound"] == 725.0 assert stats["Upper bound"] == 725.0

27
miplearn/solvers/tests/test_learning_solver.py
View File

@@ -20,11 +20,13 @@ def test_learning_solver():
for internal_solver in _get_internal_solvers(): for internal_solver in _get_internal_solvers():
logger.info("Solver: %s" % internal_solver) logger.info("Solver: %s" % internal_solver)
instance = _get_instance(internal_solver) instance = _get_instance(internal_solver)
solver = LearningSolver(time_limit=300, solver = LearningSolver(
gap_tolerance=1e-3, time_limit=300,
threads=1, gap_tolerance=1e-3,
solver=internal_solver, threads=1,
mode=mode) solver=internal_solver,
mode=mode,
)
solver.solve(instance) solver.solve(instance)
assert instance.solution["x"][0] == 1.0 assert instance.solution["x"][0] == 1.0
@@ -74,37 +76,36 @@ def test_solve_fit_from_disk():
filenames = [] filenames = []
for k in range(3): for k in range(3):
instance = _get_instance(internal_solver) instance = _get_instance(internal_solver)
with tempfile.NamedTemporaryFile(suffix=".pkl", with tempfile.NamedTemporaryFile(suffix=".pkl", delete=False) as file:
delete=False) as file:
filenames += [file.name] filenames += [file.name]
pickle.dump(instance, file) pickle.dump(instance, file)
# Test: solve # Test: solve
solver = LearningSolver(solver=internal_solver) solver = LearningSolver(solver=internal_solver)
solver.solve(filenames[0]) solver.solve(filenames[0])
with open(filenames[0], "rb") as file: with open(filenames[0], "rb") as file:
instance = pickle.load(file) instance = pickle.load(file)
assert hasattr(instance, "solution") assert hasattr(instance, "solution")
# Test: parallel_solve # Test: parallel_solve
solver.parallel_solve(filenames) solver.parallel_solve(filenames)
for filename in filenames: for filename in filenames:
with open(filename, "rb") as file: with open(filename, "rb") as file:
instance = pickle.load(file) instance = pickle.load(file)
assert hasattr(instance, "solution") assert hasattr(instance, "solution")
# Test: solve (with specified output) # Test: solve (with specified output)
output = [f + ".out" for f in filenames] output = [f + ".out" for f in filenames]
solver.solve(filenames[0], output=output[0]) solver.solve(filenames[0], output=output[0])
assert os.path.isfile(output[0]) assert os.path.isfile(output[0])
# Test: parallel_solve (with specified output) # Test: parallel_solve (with specified output)
solver.parallel_solve(filenames, output=output) solver.parallel_solve(filenames, output=output)
for filename in output: for filename in output:
assert os.path.isfile(filename) assert os.path.isfile(filename)
# Delete temporary files # Delete temporary files
for filename in filenames: for filename in filenames:
os.remove(filename) os.remove(filename)
for filename in output: for filename in output:
os.remove(filename) os.remove(filename)

12
miplearn/tests/__init__.py
View File

@@ -8,14 +8,14 @@ from miplearn.problems.knapsack import KnapsackInstance
def get_test_pyomo_instances(): def get_test_pyomo_instances():
instances = [ instances = [
KnapsackInstance( KnapsackInstance(
weights=[23., 26., 20., 18.], weights=[23.0, 26.0, 20.0, 18.0],
prices=[505., 352., 458., 220.], prices=[505.0, 352.0, 458.0, 220.0],
capacity=67., capacity=67.0,
), ),
KnapsackInstance( KnapsackInstance(
weights=[25., 30., 22., 18.], weights=[25.0, 30.0, 22.0, 18.0],
prices=[500., 365., 420., 150.], prices=[500.0, 365.0, 420.0, 150.0],
capacity=70., capacity=70.0,
), ),
] ]
models = [instance.to_model() for instance in instances] models = [instance.to_model() for instance in instances]

13
miplearn/tests/test_benchmark.py
View File

@@ -11,8 +11,9 @@ from scipy.stats import randint
def test_benchmark(): def test_benchmark():
# Generate training and test instances # Generate training and test instances
train_instances = MaxWeightStableSetGenerator(n=randint(low=25, high=26)).generate(5) generator = MaxWeightStableSetGenerator(n=randint(low=25, high=26))
test_instances = MaxWeightStableSetGenerator(n=randint(low=25, high=26)).generate(3) train_instances = generator.generate(5)
test_instances = generator.generate(3)
# Training phase... # Training phase...
training_solver = LearningSolver() training_solver = LearningSolver()
@@ -26,11 +27,11 @@ def test_benchmark():
benchmark = BenchmarkRunner(test_solvers) benchmark = BenchmarkRunner(test_solvers)
benchmark.fit(train_instances) benchmark.fit(train_instances)
benchmark.parallel_solve(test_instances, n_jobs=2, n_trials=2) benchmark.parallel_solve(test_instances, n_jobs=2, n_trials=2)
assert benchmark.raw_results().values.shape == (12,16) assert benchmark.raw_results().values.shape == (12, 16)
benchmark.save_results("/tmp/benchmark.csv") benchmark.save_results("/tmp/benchmark.csv")
assert os.path.isfile("/tmp/benchmark.csv") assert os.path.isfile("/tmp/benchmark.csv")
benchmark = BenchmarkRunner(test_solvers) benchmark = BenchmarkRunner(test_solvers)
benchmark.load_results("/tmp/benchmark.csv") benchmark.load_results("/tmp/benchmark.csv")
assert benchmark.raw_results().values.shape == (12,16) assert benchmark.raw_results().values.shape == (12, 16)

58
miplearn/tests/test_extractors.py
View File

@@ -3,25 +3,28 @@
# Released under the modified BSD license. See COPYING.md for more details. # Released under the modified BSD license. See COPYING.md for more details.
from miplearn.problems.knapsack import KnapsackInstance from miplearn.problems.knapsack import KnapsackInstance
from miplearn import (LearningSolver, from miplearn import (
SolutionExtractor, LearningSolver,
InstanceFeaturesExtractor, SolutionExtractor,
VariableFeaturesExtractor, InstanceFeaturesExtractor,
) VariableFeaturesExtractor,
)
import numpy as np import numpy as np
import pyomo.environ as pe import pyomo.environ as pe
def _get_instances(): def _get_instances():
instances = [ instances = [
KnapsackInstance(weights=[1., 2., 3.], KnapsackInstance(
prices=[10., 20., 30.], weights=[1.0, 2.0, 3.0],
capacity=2.5, prices=[10.0, 20.0, 30.0],
), capacity=2.5,
KnapsackInstance(weights=[3., 4., 5.], ),
prices=[20., 30., 40.], KnapsackInstance(
capacity=4.5, weights=[3.0, 4.0, 5.0],
), prices=[20.0, 30.0, 40.0],
capacity=4.5,
),
] ]
models = [instance.to_model() for instance in instances] models = [instance.to_model() for instance in instances]
solver = LearningSolver() solver = LearningSolver()
@@ -38,25 +41,30 @@ def test_solution_extractor():
assert isinstance(features["default"], np.ndarray) assert isinstance(features["default"], np.ndarray)
assert features["default"].shape == (6, 2) assert features["default"].shape == (6, 2)
assert features["default"].ravel().tolist() == [ assert features["default"].ravel().tolist() == [
1., 0., 1.0,
0., 1., 0.0,
1., 0., 0.0,
1., 0., 1.0,
0., 1., 1.0,
1., 0., 0.0,
1.0,
0.0,
0.0,
1.0,
1.0,
0.0,
] ]
def test_instance_features_extractor(): def test_instance_features_extractor():
instances, models = _get_instances() instances, models = _get_instances()
features = InstanceFeaturesExtractor().extract(instances) features = InstanceFeaturesExtractor().extract(instances)
assert features.shape == (2,3) assert features.shape == (2, 3)
def test_variable_features_extractor(): def test_variable_features_extractor():
instances, models = _get_instances() instances, models = _get_instances()
features = VariableFeaturesExtractor().extract(instances) features = VariableFeaturesExtractor().extract(instances)
assert isinstance(features, dict) assert isinstance(features, dict)
assert "default" in features assert "default" in features
assert features["default"].shape == (6,5) assert features["default"].shape == (6, 5)

3
pyproject.toml Normal file
View File

@@ -0,0 +1,3 @@
[tool.black]
py36 = true
include = '\.pyi?$'

2
requirements.txt
View File

@@ -12,3 +12,5 @@ python-markdown-math~=0.8
seaborn~=0.11 seaborn~=0.11
scikit-learn~=0.23 scikit-learn~=0.23
tqdm~=4.54 tqdm~=4.54
black==20.8b1
pre-commit~=2.9