Spaces:
Sleeping
Sleeping
File size: 7,944 Bytes
177c40c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 |
"""
Constraint definitions for your optimization problem.
Constraints are defined using a fluent API:
1. for_each(Entity) - iterate over all entities
2. filter(predicate) - keep only matching entities
3. join(OtherEntity, ...) - combine with other entities
4. group_by(key, collector) - aggregate by key
5. penalize(weight) or reward(weight) - affect the score
TODO: Replace these example constraints with your own business rules.
"""
from solverforge_legacy.solver.score import (
constraint_provider,
ConstraintFactory,
Joiners,
HardSoftScore,
ConstraintCollectors,
)
from .domain import Resource, Task
# =============================================================================
# CONSTRAINT WEIGHTS
# =============================================================================
# Global weights that can be adjusted at runtime via the REST API.
# Weight 0 = disabled, 100 = full strength.
# Set by rest_api.py before solving starts.
CONSTRAINT_WEIGHTS = {
'required_skill': 100, # Hard constraint
'resource_capacity': 100, # Hard constraint
'minimize_duration': 50, # Soft constraint
'balance_load': 50, # Soft constraint
}
def get_weight(name: str) -> int:
"""Get the weight for a constraint (0-100 scale)."""
return CONSTRAINT_WEIGHTS.get(name, 100)
@constraint_provider
def define_constraints(constraint_factory: ConstraintFactory):
"""
Define all constraints for the optimization problem.
Returns a list of constraints, evaluated in order:
- Hard constraints: Must be satisfied (score < 0 = infeasible)
- Soft constraints: Should be optimized (higher = better)
"""
return [
# Hard constraints (must be satisfied)
required_skill(constraint_factory),
resource_capacity(constraint_factory),
# Soft constraints (optimize these)
minimize_total_duration(constraint_factory),
balance_resource_load(constraint_factory),
]
# =============================================================================
# HARD CONSTRAINTS
# =============================================================================
def required_skill(constraint_factory: ConstraintFactory):
"""
Hard: Each task must be assigned to a resource with the required skill.
Pattern: for_each -> filter -> penalize
NOTE: We check task.resource is not None FIRST, because unassigned tasks
should not be penalized - they're just not yet assigned.
NOTE: We use len(str(task.required_skill)) > 0 instead of just task.required_skill
because the value may be a Java String object which doesn't work with Python's
boolean operators directly.
WEIGHT: When weight=0, this constraint is effectively disabled.
"""
weight = get_weight('required_skill')
if weight == 0:
# Return a no-op constraint when disabled
return (
constraint_factory.for_each(Task)
.filter(lambda task: False) # Never matches
.penalize(HardSoftScore.ONE_HARD)
.as_constraint("Required skill missing")
)
return (
constraint_factory.for_each(Task)
.filter(lambda task: task.resource is not None
and len(str(task.required_skill)) > 0
and not task.has_required_skill())
.penalize(HardSoftScore.ONE_HARD, lambda task: weight)
.as_constraint("Required skill missing")
)
def resource_capacity(constraint_factory: ConstraintFactory):
"""
Hard: Total task duration per resource must not exceed capacity.
Pattern: for_each -> group_by -> filter -> penalize
WEIGHT: When weight=0, this constraint is effectively disabled.
"""
weight = get_weight('resource_capacity')
if weight == 0:
return (
constraint_factory.for_each(Task)
.filter(lambda task: False)
.penalize(HardSoftScore.ONE_HARD)
.as_constraint("Resource capacity exceeded")
)
return (
constraint_factory.for_each(Task)
.group_by(
lambda task: task.resource,
ConstraintCollectors.sum(lambda task: task.duration)
)
.filter(lambda resource, total_duration:
resource is not None and total_duration > resource.capacity)
.penalize(
HardSoftScore.ONE_HARD,
lambda resource, total_duration: (total_duration - resource.capacity) * weight // 100
)
.as_constraint("Resource capacity exceeded")
)
# =============================================================================
# SOFT CONSTRAINTS
# =============================================================================
def minimize_total_duration(constraint_factory: ConstraintFactory):
"""
Soft: Prefer shorter total duration (makespan).
Pattern: for_each -> penalize with weight function
WEIGHT: Penalty multiplied by weight/100.
"""
weight = get_weight('minimize_duration')
if weight == 0:
return (
constraint_factory.for_each(Task)
.filter(lambda task: False)
.penalize(HardSoftScore.ONE_SOFT)
.as_constraint("Minimize total duration")
)
return (
constraint_factory.for_each(Task)
.filter(lambda task: task.resource is not None)
.penalize(HardSoftScore.ONE_SOFT, lambda task: task.duration * weight // 100)
.as_constraint("Minimize total duration")
)
def balance_resource_load(constraint_factory: ConstraintFactory):
"""
Soft: Balance workload fairly across all resources.
Pattern: for_each -> group_by -> complement -> group_by(loadBalance) -> penalize
WEIGHT: Penalty multiplied by weight/100.
"""
weight = get_weight('balance_load')
if weight == 0:
return (
constraint_factory.for_each(Task)
.filter(lambda task: False)
.penalize(HardSoftScore.ONE_SOFT)
.as_constraint("Balance resource load")
)
return (
constraint_factory.for_each(Task)
.group_by(
lambda task: task.resource,
ConstraintCollectors.sum(lambda task: task.duration)
)
.complement(Resource, lambda r: 0) # Include resources with 0 tasks
.group_by(
ConstraintCollectors.load_balance(
lambda resource, duration: resource,
lambda resource, duration: duration,
)
)
.penalize(
HardSoftScore.ONE_SOFT,
lambda load_balance: int(load_balance.unfairness()) * weight // 100
)
.as_constraint("Balance resource load")
)
# =============================================================================
# ADDITIONAL CONSTRAINT PATTERNS (commented examples)
# =============================================================================
# def no_overlapping_tasks(constraint_factory: ConstraintFactory):
# """
# Example: Two tasks on same resource cannot overlap in time.
#
# Pattern: for_each_unique_pair with Joiners
# """
# return (
# constraint_factory.for_each_unique_pair(
# Task,
# Joiners.equal(lambda task: task.resource),
# Joiners.overlapping(
# lambda task: task.start_time,
# lambda task: task.end_time
# ),
# )
# .penalize(HardSoftScore.ONE_HARD)
# .as_constraint("Overlapping tasks")
# )
# def preferred_resource(constraint_factory: ConstraintFactory):
# """
# Example: Reward tasks assigned to their preferred resource.
#
# Pattern: for_each -> filter -> reward
# """
# return (
# constraint_factory.for_each(Task)
# .filter(lambda task: task.resource == task.preferred_resource)
# .reward(HardSoftScore.ONE_SOFT)
# .as_constraint("Preferred resource")
# )
|