You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
592 lines
20 KiB
592 lines
20 KiB
# MIPLearn: Extensible Framework for Learning-Enhanced Mixed-Integer Optimization
|
|
# Copyright (C) 2020-2021, UChicago Argonne, LLC. All rights reserved.
|
|
# Released under the modified BSD license. See COPYING.md for more details.
|
|
|
|
import logging
|
|
import time
|
|
import traceback
|
|
from typing import Optional, List, Any, cast, Dict, Tuple, Callable, IO, Union
|
|
|
|
from overrides import overrides
|
|
from p_tqdm import p_map, p_umap
|
|
from tqdm.auto import tqdm
|
|
|
|
from miplearn.features.sample import Hdf5Sample, Sample
|
|
from miplearn.components.component import Component
|
|
from miplearn.components.dynamic_lazy import DynamicLazyConstraintsComponent
|
|
from miplearn.components.dynamic_user_cuts import UserCutsComponent
|
|
from miplearn.components.objective import ObjectiveValueComponent
|
|
from miplearn.components.primal import PrimalSolutionComponent
|
|
from miplearn.features.extractor import FeaturesExtractor
|
|
from miplearn.instance.base import Instance
|
|
from miplearn.solvers import _RedirectOutput
|
|
from miplearn.solvers.internal import InternalSolver
|
|
from miplearn.solvers.pyomo.gurobi import GurobiPyomoSolver
|
|
from miplearn.types import LearningSolveStats, ConstraintName
|
|
import gzip
|
|
import pickle
|
|
import miplearn
|
|
import json
|
|
from os.path import exists
|
|
from os import remove
|
|
import pyomo.environ as pe
|
|
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
class PyomoFindLazyCutCallbackHandler:
|
|
def __init__(self):
|
|
pass
|
|
|
|
def value(self, var):
|
|
return var.value
|
|
|
|
|
|
class PyomoEnforceLazyCutsCallbackHandler:
|
|
def __init__(self, opt, model):
|
|
self.model = model
|
|
self.opt = opt
|
|
if not hasattr(model, "miplearn_lazy_cb"):
|
|
model.miplearn_lazy_cb = pe.ConstraintList()
|
|
|
|
def enforce(self, expr):
|
|
constr = self.model.miplearn_lazy_cb.add(expr=expr)
|
|
self.opt.add_constraint(constr)
|
|
|
|
|
|
class FileInstanceWrapper(Instance):
|
|
def __init__(
|
|
self, data_filename: Any, build_model: Callable, mode: Optional[str] = None
|
|
):
|
|
super().__init__()
|
|
assert data_filename.endswith(".pkl.gz")
|
|
self.filename = data_filename
|
|
self.sample_filename = data_filename.replace(".pkl.gz", ".h5")
|
|
self.build_model = build_model
|
|
self.mode = mode
|
|
self.sample = None
|
|
self.model = None
|
|
|
|
@overrides
|
|
def to_model(self) -> Any:
|
|
if self.model is None:
|
|
self.model = miplearn.load(self.filename, self.build_model)
|
|
return self.model
|
|
|
|
@overrides
|
|
def create_sample(self) -> Sample:
|
|
return self.sample
|
|
|
|
@overrides
|
|
def get_samples(self) -> List[Sample]:
|
|
return [self.sample]
|
|
|
|
@overrides
|
|
def free(self) -> None:
|
|
self.sample.file.close()
|
|
|
|
@overrides
|
|
def load(self) -> None:
|
|
if self.mode is None:
|
|
self.mode = "r+" if exists(self.sample_filename) else "w"
|
|
self.sample = Hdf5Sample(self.sample_filename, mode=self.mode)
|
|
|
|
@overrides
|
|
def has_dynamic_lazy_constraints(self) -> bool:
|
|
assert hasattr(self, "model")
|
|
return hasattr(self.model, "_miplearn_find_lazy_cuts")
|
|
|
|
@overrides
|
|
def find_violated_lazy_constraints(
|
|
self,
|
|
solver: "InternalSolver",
|
|
model: Any,
|
|
) -> Dict[ConstraintName, Any]:
|
|
if not hasattr(self.model, "_miplearn_find_lazy_cuts"):
|
|
return {}
|
|
cb = PyomoFindLazyCutCallbackHandler()
|
|
violations = model._miplearn_find_lazy_cuts(cb)
|
|
return {json.dumps(v).encode(): v for v in violations}
|
|
|
|
@overrides
|
|
def enforce_lazy_constraint(
|
|
self,
|
|
solver: "InternalSolver",
|
|
model: Any,
|
|
violation: Any,
|
|
) -> None:
|
|
assert isinstance(solver, GurobiPyomoSolver)
|
|
cb = PyomoEnforceLazyCutsCallbackHandler(solver._pyomo_solver, model)
|
|
model._miplearn_enforce_lazy_cuts(cb, violation)
|
|
|
|
|
|
class MemoryInstanceWrapper(Instance):
|
|
def __init__(self, model: Any) -> None:
|
|
super().__init__()
|
|
assert model is not None
|
|
self.model = model
|
|
|
|
@overrides
|
|
def to_model(self) -> Any:
|
|
return self.model
|
|
|
|
@overrides
|
|
def has_dynamic_lazy_constraints(self) -> bool:
|
|
assert hasattr(self, "model")
|
|
return hasattr(self.model, "_miplearn_find_lazy_cuts")
|
|
|
|
@overrides
|
|
def find_violated_lazy_constraints(
|
|
self,
|
|
solver: "InternalSolver",
|
|
model: Any,
|
|
) -> Dict[ConstraintName, Any]:
|
|
cb = PyomoFindLazyCutCallbackHandler()
|
|
violations = model._miplearn_find_lazy_cuts(cb)
|
|
return {json.dumps(v).encode(): v for v in violations}
|
|
|
|
@overrides
|
|
def enforce_lazy_constraint(
|
|
self,
|
|
solver: "InternalSolver",
|
|
model: Any,
|
|
violation: Any,
|
|
) -> None:
|
|
assert isinstance(solver, GurobiPyomoSolver)
|
|
cb = PyomoEnforceLazyCutsCallbackHandler(solver._pyomo_solver, model)
|
|
model._miplearn_enforce_lazy_cuts(cb, violation)
|
|
|
|
|
|
class _GlobalVariables:
|
|
def __init__(self) -> None:
|
|
self.solver: Optional[LearningSolver] = None
|
|
self.build_model: Optional[Callable] = None
|
|
self.filenames: Optional[List[str]] = None
|
|
self.skip = False
|
|
|
|
|
|
# Global variables used for multiprocessing. Global variables are copied by the
|
|
# operating system when the process forks. Local variables are copied through
|
|
# serialization, which is a much slower process.
|
|
_GLOBAL = [_GlobalVariables()]
|
|
|
|
|
|
def _parallel_solve(
|
|
idx: int,
|
|
) -> Tuple[Optional[int], Optional[LearningSolveStats]]:
|
|
solver = _GLOBAL[0].solver
|
|
filenames = _GLOBAL[0].filenames
|
|
build_model = _GLOBAL[0].build_model
|
|
skip = _GLOBAL[0].skip
|
|
assert solver is not None
|
|
try:
|
|
stats = solver.solve([filenames[idx]], build_model, skip=skip)
|
|
return idx, stats[0]
|
|
except Exception as e:
|
|
traceback.print_exc()
|
|
logger.exception(f"Exception while solving {filenames[idx]}. Ignoring.")
|
|
return idx, None
|
|
|
|
|
|
class LearningSolver:
|
|
"""
|
|
Mixed-Integer Linear Programming (MIP) solver that extracts information
|
|
from previous runs and uses Machine Learning methods to accelerate the
|
|
solution of new (yet unseen) instances.
|
|
|
|
Parameters
|
|
----------
|
|
components: List[Component]
|
|
Set of components in the solver. By default, includes
|
|
`ObjectiveValueComponent`, `PrimalSolutionComponent`,
|
|
`DynamicLazyConstraintsComponent` and `UserCutsComponent`.
|
|
mode: str
|
|
If "exact", solves problem to optimality, keeping all optimality
|
|
guarantees provided by the MIP solver. If "heuristic", uses machine
|
|
learning more aggressively, and may return suboptimal solutions.
|
|
solver: Callable[[], InternalSolver]
|
|
A callable that constructs the internal solver. If None is provided,
|
|
use GurobiPyomoSolver.
|
|
use_lazy_cb: bool
|
|
If true, use native solver callbacks for enforcing lazy constraints,
|
|
instead of a simple loop. May not be supported by all solvers.
|
|
solve_lp: bool
|
|
If true, solve the root LP relaxation before solving the MIP. This
|
|
option should be activated if the LP relaxation is not very
|
|
expensive to solve and if it provides good hints for the integer
|
|
solution.
|
|
"""
|
|
|
|
def __init__(
|
|
self,
|
|
components: Optional[List[Component]] = None,
|
|
mode: str = "exact",
|
|
solver: Optional[InternalSolver] = None,
|
|
use_lazy_cb: bool = False,
|
|
solve_lp: bool = True,
|
|
extractor: Optional[FeaturesExtractor] = None,
|
|
extract_lhs: bool = True,
|
|
extract_sa: bool = True,
|
|
) -> None:
|
|
if solver is None:
|
|
solver = GurobiPyomoSolver()
|
|
if extractor is None:
|
|
extractor = FeaturesExtractor(
|
|
with_sa=extract_sa,
|
|
with_lhs=extract_lhs,
|
|
)
|
|
assert isinstance(solver, InternalSolver)
|
|
self.components: Dict[str, Component] = {}
|
|
self.internal_solver: Optional[InternalSolver] = None
|
|
self.internal_solver_prototype: InternalSolver = solver
|
|
self.mode: str = mode
|
|
self.solve_lp: bool = solve_lp
|
|
self.tee = False
|
|
self.use_lazy_cb: bool = use_lazy_cb
|
|
self.extractor = extractor
|
|
if components is not None:
|
|
for comp in components:
|
|
self._add_component(comp)
|
|
else:
|
|
self._add_component(ObjectiveValueComponent())
|
|
self._add_component(PrimalSolutionComponent(mode=mode))
|
|
self._add_component(DynamicLazyConstraintsComponent())
|
|
self._add_component(UserCutsComponent())
|
|
assert self.mode in ["exact", "heuristic"]
|
|
|
|
def _solve(
|
|
self,
|
|
instance: Instance,
|
|
model: Any = None,
|
|
discard_output: bool = False,
|
|
tee: bool = False,
|
|
) -> LearningSolveStats:
|
|
"""
|
|
Solves the given instance. If trained machine-learning models are
|
|
available, they will be used to accelerate the solution process.
|
|
|
|
The argument `instance` may be either an Instance object or a
|
|
filename pointing to a pickled Instance object.
|
|
|
|
This method adds a new training sample to `instance.training_sample`.
|
|
If a filename is provided, then the file is modified in-place. That is,
|
|
the original file is overwritten.
|
|
|
|
If `solver.solve_lp_first` is False, the properties lp_solution and
|
|
lp_value will be set to dummy values.
|
|
|
|
Parameters
|
|
----------
|
|
instance: Instance
|
|
The instance to be solved.
|
|
model: Any
|
|
The corresponding Pyomo model. If not provided, it will be created.
|
|
discard_output: bool
|
|
If True, do not write the modified instances anywhere; simply discard
|
|
them. Useful during benchmarking.
|
|
tee: bool
|
|
If true, prints solver log to screen.
|
|
|
|
Returns
|
|
-------
|
|
LearningSolveStats
|
|
A dictionary of solver statistics containing at least the following
|
|
keys: "Lower bound", "Upper bound", "Wallclock time", "Nodes",
|
|
"Sense", "Log", "Warm start value" and "LP value".
|
|
|
|
Additional components may generate additional keys. For example,
|
|
ObjectiveValueComponent adds the keys "Predicted LB" and
|
|
"Predicted UB". See the documentation of each component for more
|
|
details.
|
|
"""
|
|
|
|
# Generate model
|
|
# -------------------------------------------------------
|
|
instance.load()
|
|
if model is None:
|
|
with _RedirectOutput([]):
|
|
model = instance.to_model()
|
|
|
|
# Initialize training sample
|
|
# -------------------------------------------------------
|
|
sample = instance.create_sample()
|
|
|
|
# Initialize stats
|
|
# -------------------------------------------------------
|
|
stats: LearningSolveStats = {}
|
|
|
|
# Initialize internal solver
|
|
# -------------------------------------------------------
|
|
self.tee = tee
|
|
self.internal_solver = self.internal_solver_prototype.clone()
|
|
assert self.internal_solver is not None
|
|
assert isinstance(self.internal_solver, InternalSolver)
|
|
self.internal_solver.set_instance(instance, model)
|
|
|
|
# Extract features (after-load)
|
|
# -------------------------------------------------------
|
|
logger.info("Extracting features (after-load)...")
|
|
initial_time = time.time()
|
|
self.extractor.extract_after_load_features(
|
|
instance, self.internal_solver, sample
|
|
)
|
|
logger.info(
|
|
"Features (after-load) extracted in %.2f seconds"
|
|
% (time.time() - initial_time)
|
|
)
|
|
|
|
callback_args = (
|
|
self,
|
|
instance,
|
|
model,
|
|
stats,
|
|
sample,
|
|
)
|
|
|
|
# Solve root LP relaxation
|
|
# -------------------------------------------------------
|
|
lp_stats = None
|
|
if self.solve_lp:
|
|
logger.debug("Running before_solve_lp callbacks...")
|
|
for component in self.components.values():
|
|
component.before_solve_lp(*callback_args)
|
|
|
|
logger.info("Solving root LP relaxation...")
|
|
lp_stats = self.internal_solver.solve_lp(tee=tee)
|
|
stats.update(cast(LearningSolveStats, lp_stats.__dict__))
|
|
assert lp_stats.lp_wallclock_time is not None
|
|
logger.info(
|
|
"LP relaxation solved in %.2f seconds" % lp_stats.lp_wallclock_time
|
|
)
|
|
|
|
logger.debug("Running after_solve_lp callbacks...")
|
|
for component in self.components.values():
|
|
component.after_solve_lp(*callback_args)
|
|
|
|
# Extract features (after-lp)
|
|
# -------------------------------------------------------
|
|
logger.info("Extracting features (after-lp)...")
|
|
initial_time = time.time()
|
|
self.extractor.extract_after_lp_features(
|
|
self.internal_solver, sample, lp_stats
|
|
)
|
|
logger.info(
|
|
"Features (after-lp) extracted in %.2f seconds"
|
|
% (time.time() - initial_time)
|
|
)
|
|
|
|
# Callback wrappers
|
|
# -------------------------------------------------------
|
|
def iteration_cb_wrapper() -> bool:
|
|
should_repeat = False
|
|
for comp in self.components.values():
|
|
if comp.iteration_cb(self, instance, model):
|
|
should_repeat = True
|
|
return should_repeat
|
|
|
|
def lazy_cb_wrapper(
|
|
cb_solver: InternalSolver,
|
|
cb_model: Any,
|
|
) -> None:
|
|
for comp in self.components.values():
|
|
comp.lazy_cb(self, instance, model)
|
|
|
|
def user_cut_cb_wrapper(
|
|
cb_solver: InternalSolver,
|
|
cb_model: Any,
|
|
) -> None:
|
|
for comp in self.components.values():
|
|
comp.user_cut_cb(self, instance, model)
|
|
|
|
lazy_cb = None
|
|
if self.use_lazy_cb:
|
|
lazy_cb = lazy_cb_wrapper
|
|
|
|
user_cut_cb = None
|
|
if instance.has_user_cuts():
|
|
user_cut_cb = user_cut_cb_wrapper
|
|
|
|
# Before-solve callbacks
|
|
# -------------------------------------------------------
|
|
logger.debug("Running before_solve_mip callbacks...")
|
|
for component in self.components.values():
|
|
component.before_solve_mip(*callback_args)
|
|
|
|
# Solve MIP
|
|
# -------------------------------------------------------
|
|
logger.info("Solving MIP...")
|
|
mip_stats = self.internal_solver.solve(
|
|
tee=tee,
|
|
iteration_cb=iteration_cb_wrapper,
|
|
user_cut_cb=user_cut_cb,
|
|
lazy_cb=lazy_cb,
|
|
)
|
|
assert mip_stats.mip_wallclock_time is not None
|
|
logger.info("MIP solved in %.2f seconds" % mip_stats.mip_wallclock_time)
|
|
stats.update(cast(LearningSolveStats, mip_stats.__dict__))
|
|
stats["Solver"] = "default"
|
|
stats["Gap"] = self._compute_gap(
|
|
ub=mip_stats.mip_upper_bound,
|
|
lb=mip_stats.mip_lower_bound,
|
|
)
|
|
stats["Mode"] = self.mode
|
|
|
|
# Extract features (after-mip)
|
|
# -------------------------------------------------------
|
|
logger.info("Extracting features (after-mip)...")
|
|
initial_time = time.time()
|
|
for (k, v) in mip_stats.__dict__.items():
|
|
sample.put_scalar(k, v)
|
|
self.extractor.extract_after_mip_features(self.internal_solver, sample)
|
|
logger.info(
|
|
"Features (after-mip) extracted in %.2f seconds"
|
|
% (time.time() - initial_time)
|
|
)
|
|
|
|
# After-solve callbacks
|
|
# -------------------------------------------------------
|
|
logger.debug("Calling after_solve_mip callbacks...")
|
|
for component in self.components.values():
|
|
component.after_solve_mip(*callback_args)
|
|
|
|
# Flush
|
|
# -------------------------------------------------------
|
|
if not discard_output:
|
|
instance.flush()
|
|
instance.free()
|
|
|
|
return stats
|
|
|
|
def solve(
|
|
self,
|
|
arg: Union[Any, List[str]],
|
|
build_model: Optional[Callable] = None,
|
|
tee: bool = False,
|
|
progress: bool = False,
|
|
skip: bool = False,
|
|
) -> Union[LearningSolveStats, List[LearningSolveStats]]:
|
|
if isinstance(arg, list):
|
|
assert build_model is not None
|
|
stats = []
|
|
for i in tqdm(arg, disable=not progress):
|
|
instance = FileInstanceWrapper(i, build_model)
|
|
solved = False
|
|
if exists(instance.sample_filename):
|
|
try:
|
|
with Hdf5Sample(instance.sample_filename, mode="r") as sample:
|
|
if sample.get_scalar("mip_lower_bound"):
|
|
solved = True
|
|
except OSError:
|
|
# File exists but it is unreadable/corrupted. Delete it.
|
|
remove(instance.sample_filename)
|
|
if solved and skip:
|
|
stats.append({})
|
|
else:
|
|
s = self._solve(instance, tee=tee)
|
|
|
|
# Export to gzipped MPS file
|
|
mps_filename = instance.sample_filename.replace(".h5", ".mps")
|
|
instance.model.write(
|
|
filename=mps_filename,
|
|
io_options={
|
|
"labeler": pe.NameLabeler(),
|
|
"skip_objective_sense": True,
|
|
},
|
|
)
|
|
with open(mps_filename, "rb") as original:
|
|
with gzip.open(f"{mps_filename}.gz", "wb") as compressed:
|
|
compressed.writelines(original)
|
|
remove(mps_filename)
|
|
|
|
stats.append(s)
|
|
return stats
|
|
else:
|
|
return self._solve(MemoryInstanceWrapper(arg), tee=tee)
|
|
|
|
def fit(
|
|
self,
|
|
filenames: List[str],
|
|
build_model: Callable,
|
|
progress: bool = False,
|
|
n_jobs: int = 1,
|
|
) -> None:
|
|
instances: List[Instance] = [
|
|
FileInstanceWrapper(f, build_model, mode="r") for f in filenames
|
|
]
|
|
self._fit(instances, progress=progress, n_jobs=n_jobs)
|
|
|
|
def parallel_solve(
|
|
self,
|
|
filenames: List[str],
|
|
build_model: Optional[Callable] = None,
|
|
n_jobs: int = 4,
|
|
progress: bool = False,
|
|
label: str = "solve",
|
|
skip: bool = False,
|
|
) -> List[LearningSolveStats]:
|
|
self.internal_solver = None
|
|
self._silence_miplearn_logger()
|
|
_GLOBAL[0].solver = self
|
|
_GLOBAL[0].build_model = build_model
|
|
_GLOBAL[0].filenames = filenames
|
|
_GLOBAL[0].skip = skip
|
|
results = p_umap(
|
|
_parallel_solve,
|
|
list(range(len(filenames))),
|
|
num_cpus=n_jobs,
|
|
disable=not progress,
|
|
desc=label,
|
|
)
|
|
stats: List[LearningSolveStats] = [{} for _ in range(len(filenames))]
|
|
for (idx, s) in results:
|
|
if s:
|
|
stats[idx] = s
|
|
self._restore_miplearn_logger()
|
|
return stats
|
|
|
|
def _fit(
|
|
self,
|
|
training_instances: List[Instance],
|
|
n_jobs: int = 1,
|
|
progress: bool = False,
|
|
) -> None:
|
|
if len(training_instances) == 0:
|
|
logger.warning("Empty list of training instances provided. Skipping.")
|
|
return
|
|
Component.fit_multiple(
|
|
list(self.components.values()),
|
|
training_instances,
|
|
n_jobs=n_jobs,
|
|
progress=progress,
|
|
)
|
|
|
|
def _add_component(self, component: Component) -> None:
|
|
name = component.__class__.__name__
|
|
self.components[name] = component
|
|
|
|
def _silence_miplearn_logger(self) -> None:
|
|
miplearn_logger = logging.getLogger("miplearn")
|
|
self.prev_log_level = miplearn_logger.getEffectiveLevel()
|
|
miplearn_logger.setLevel(logging.WARNING)
|
|
|
|
def _restore_miplearn_logger(self) -> None:
|
|
miplearn_logger = logging.getLogger("miplearn")
|
|
miplearn_logger.setLevel(self.prev_log_level)
|
|
|
|
def __getstate__(self) -> Dict:
|
|
self.internal_solver = None
|
|
return self.__dict__
|
|
|
|
@staticmethod
|
|
def _compute_gap(ub: Optional[float], lb: Optional[float]) -> Optional[float]:
|
|
if lb is None or ub is None or lb * ub < 0:
|
|
# solver did not find a solution and/or bound
|
|
return None
|
|
elif abs(ub - lb) < 1e-6:
|
|
# avoid division by zero when ub = lb = 0
|
|
return 0.0
|
|
else:
|
|
# divide by max(abs(ub),abs(lb)) to ensure gap <= 1
|
|
return (ub - lb) / max(abs(ub), abs(lb))
|