Source code for cvxpy.problems.objective
"""
Copyright 2013 Steven Diamond
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""
import cvxpy.lin_ops.lin_utils as lu
import cvxpy.utilities as u
from cvxpy.error import DCPError
from cvxpy.expressions.expression import Expression
from cvxpy.interface.matrix_utilities import scalar_value
from cvxpy.utilities import scopes
class Objective(u.Canonical):
"""An optimization objective.
Parameters
----------
expr : Expression
The expression to act upon. Must be a scalar.
Raises
------
ValueError
If expr is not a scalar.
"""
NAME = "objective"
def __init__(self, expr) -> None:
self.args = [Expression.cast_to_const(expr)]
# Validate that the objective resolves to a scalar.
if not self.args[0].is_scalar():
raise ValueError("The '%s' objective must resolve to a scalar."
% self.NAME)
if not self.args[0].is_real():
raise ValueError("The '%s' objective must be real valued."
% self.NAME)
def __repr__(self) -> str:
return "%s(%s)" % (self.__class__.__name__, repr(self.args[0]))
def __str__(self) -> str:
return ' '.join([self.NAME, self.args[0].name()])
def __radd__(self, other):
if other == 0:
return self
else:
raise NotImplementedError()
def __sub__(self, other):
if not isinstance(other, (Minimize, Maximize)):
raise NotImplementedError()
# Objectives must opposites
return self + (-other)
def __rsub__(self, other):
if other == 0:
return -self
else:
raise NotImplementedError()
def __mul__(self, other):
if not isinstance(other, (int, float)):
raise NotImplementedError()
# If negative, reverse the direction of objective
if (type(self) == Maximize) == (other < 0.0):
return Minimize(self.args[0] * other)
else:
return Maximize(self.args[0] * other)
__rmul__ = __mul__
def __div__(self, other):
if not isinstance(other, (int, float)):
raise NotImplementedError()
return self * (1.0/other)
__truediv__ = __div__
@property
def value(self):
"""The value of the objective expression.
"""
v = self.args[0].value
if v is None:
return None
else:
return scalar_value(v)
def is_quadratic(self) -> bool:
"""Returns if the objective is a quadratic function.
"""
return self.args[0].is_quadratic()
def is_qpwa(self) -> bool:
"""Returns if the objective is a quadratic of piecewise affine.
"""
return self.args[0].is_qpwa()
[docs]class Minimize(Objective):
"""An optimization objective for minimization.
Parameters
----------
expr : Expression
The expression to minimize. Must be a scalar.
Raises
------
ValueError
If expr is not a scalar.
"""
NAME = "minimize"
def __neg__(self) -> "Maximize":
return Maximize(-self.args[0])
def __add__(self, other):
if not isinstance(other, (Minimize, Maximize)):
raise NotImplementedError()
# Objectives must both be Minimize.
if type(other) is Minimize:
return Minimize(self.args[0] + other.args[0])
else:
raise DCPError("Problem does not follow DCP rules.")
def canonicalize(self):
"""Pass on the target expression's objective and constraints.
"""
return self.args[0].canonical_form
[docs] def is_dcp(self, dpp: bool = False) -> bool:
"""The objective must be convex.
"""
if dpp:
with scopes.dpp_scope():
return self.args[0].is_convex()
return self.args[0].is_convex()
[docs] def is_dgp(self, dpp: bool = False) -> bool:
"""The objective must be log-log convex.
"""
if dpp:
with scopes.dpp_scope():
return self.args[0].is_log_log_convex()
return self.args[0].is_log_log_convex()
def is_dpp(self, context='dcp') -> bool:
with scopes.dpp_scope():
if context.lower() == 'dcp':
return self.is_dcp(dpp=True)
elif context.lower() == 'dgp':
return self.is_dgp(dpp=True)
else:
raise ValueError("Unsupported context ", context)
def is_dqcp(self) -> bool:
"""The objective must be quasiconvex.
"""
return self.args[0].is_quasiconvex()
@staticmethod
def primal_to_result(result):
"""The value of the objective given the solver primal value.
"""
return result
[docs]class Maximize(Objective):
"""An optimization objective for maximization.
Parameters
----------
expr : Expression
The expression to maximize. Must be a scalar.
Raises
------
ValueError
If expr is not a scalar.
"""
NAME = "maximize"
def __neg__(self) -> Minimize:
return Minimize(-self.args[0])
def __add__(self, other):
if not isinstance(other, (Minimize, Maximize)):
raise NotImplementedError()
# Objectives must both be Maximize.
if type(other) is Maximize:
return Maximize(self.args[0] + other.args[0])
else:
raise Exception("Problem does not follow DCP rules.")
def canonicalize(self):
"""Negates the target expression's objective.
"""
obj, constraints = self.args[0].canonical_form
return (lu.neg_expr(obj), constraints)
[docs] def is_dcp(self, dpp: bool = False) -> bool:
"""The objective must be concave.
"""
if dpp:
with scopes.dpp_scope():
return self.args[0].is_concave()
return self.args[0].is_concave()
[docs] def is_dgp(self, dpp: bool = False) -> bool:
"""The objective must be log-log concave.
"""
if dpp:
with scopes.dpp_scope():
return self.args[0].is_log_log_concave()
return self.args[0].is_log_log_concave()
def is_dpp(self, context='dcp') -> bool:
with scopes.dpp_scope():
if context.lower() == 'dcp':
return self.is_dcp(dpp=True)
elif context.lower() == 'dgp':
return self.is_dgp(dpp=True)
else:
raise ValueError("Unsupported context ", context)
def is_dqcp(self) -> bool:
"""The objective must be quasiconcave.
"""
return self.args[0].is_quasiconcave()
@staticmethod
def primal_to_result(result):
"""The value of the objective given the solver primal value.
"""
return -result