public final class PrimalDualHybridGradientParams extends com.google.protobuf.GeneratedMessage implements PrimalDualHybridGradientParamsOrBuilder
Parameters for PrimalDualHybridGradient() in primal_dual_hybrid_gradient.h. While the defaults are generally good, it is usually worthwhile to perform a parameter sweep to find good settings for a particular family of problems. The following parameters should be considered for tuning: - restart_strategy (jointly with major_iteration_frequency) - primal_weight_update_smoothing (jointly with initial_primal_weight) - presolve_options.use_glop - l_inf_ruiz_iterations - l2_norm_rescaling In addition, tune num_threads to speed up the solve.Protobuf type
operations_research.pdlp.PrimalDualHybridGradientParams| Modifier and Type | Class and Description |
|---|---|
static class |
PrimalDualHybridGradientParams.Builder
Parameters for PrimalDualHybridGradient() in primal_dual_hybrid_gradient.h.
|
static class |
PrimalDualHybridGradientParams.LinesearchRule
Protobuf enum
operations_research.pdlp.PrimalDualHybridGradientParams.LinesearchRule |
static class |
PrimalDualHybridGradientParams.PresolveOptions
Protobuf type
operations_research.pdlp.PrimalDualHybridGradientParams.PresolveOptions |
static interface |
PrimalDualHybridGradientParams.PresolveOptionsOrBuilder |
static class |
PrimalDualHybridGradientParams.RestartStrategy
Protobuf enum
operations_research.pdlp.PrimalDualHybridGradientParams.RestartStrategy |
com.google.protobuf.GeneratedMessage.ExtendableBuilder<MessageT extends com.google.protobuf.GeneratedMessage.ExtendableMessage<MessageT>,BuilderT extends com.google.protobuf.GeneratedMessage.ExtendableBuilder<MessageT,BuilderT>>, com.google.protobuf.GeneratedMessage.ExtendableMessage<MessageT extends com.google.protobuf.GeneratedMessage.ExtendableMessage<MessageT>>, com.google.protobuf.GeneratedMessage.ExtendableMessageOrBuilder<MessageT extends com.google.protobuf.GeneratedMessage.ExtendableMessage<MessageT>>, com.google.protobuf.GeneratedMessage.FieldAccessorTable, com.google.protobuf.GeneratedMessage.GeneratedExtension<ContainingT extends com.google.protobuf.Message,T>, com.google.protobuf.GeneratedMessage.UnusedPrivateParameter| Modifier and Type | Method and Description |
|---|---|
boolean |
equals(java.lang.Object obj) |
AdaptiveLinesearchParams |
getAdaptiveLinesearchParameters()
optional .operations_research.pdlp.AdaptiveLinesearchParams adaptive_linesearch_parameters = 18; |
AdaptiveLinesearchParamsOrBuilder |
getAdaptiveLinesearchParametersOrBuilder()
optional .operations_research.pdlp.AdaptiveLinesearchParams adaptive_linesearch_parameters = 18; |
static PrimalDualHybridGradientParams |
getDefaultInstance() |
PrimalDualHybridGradientParams |
getDefaultInstanceForType() |
static com.google.protobuf.Descriptors.Descriptor |
getDescriptor() |
double |
getDiagonalQpTrustRegionSolverTolerance()
The solve tolerance of the experimental trust region solver for diagonal
QPs, controlling the accuracy of binary search over a one-dimensional
scaling parameter.
|
boolean |
getHandleSomePrimalGradientsOnFiniteBoundsAsResiduals()
See
https://developers.google.com/optimization/lp/pdlp_math#treating_some_variable_bounds_as_infinite
for a description of this flag.
|
double |
getInfiniteConstraintBoundThreshold()
Constraint bounds with absolute value at least this threshold are replaced
with infinities.
|
double |
getInitialPrimalWeight()
The initial value of the primal weight (i.e., the ratio of primal and dual
step sizes).
|
double |
getInitialStepSizeScaling()
Scaling factor applied to the initial step size (all step sizes if
linesearch_rule == CONSTANT_STEP_SIZE_RULE).
|
boolean |
getL2NormRescaling()
If true, applies L_2 norm rescaling after the Ruiz rescaling.
|
PrimalDualHybridGradientParams.LinesearchRule |
getLinesearchRule()
Linesearch rule applied at each major iteration.
|
int |
getLInfRuizIterations()
Number of L_infinity Ruiz rescaling iterations to apply to the constraint
matrix.
|
double |
getLogIntervalSeconds()
Time between iteration-level statistics logging (if `verbosity_level > 1`).
|
int |
getMajorIterationFrequency()
The frequency at which extra work is performed to make major algorithmic
decisions, e.g., performing restarts and updating the primal weight.
|
MalitskyPockParams |
getMalitskyPockParameters()
optional .operations_research.pdlp.MalitskyPockParams malitsky_pock_parameters = 19; |
MalitskyPockParamsOrBuilder |
getMalitskyPockParametersOrBuilder()
optional .operations_research.pdlp.MalitskyPockParams malitsky_pock_parameters = 19; |
double |
getNecessaryReductionForRestart()
For ADAPTIVE_HEURISTIC only: A relative reduction in the potential function
by this amount triggers a restart if, additionally, the quality of the
iterates appears to be getting worse.
|
int |
getNumShards()
For more efficient parallel computation, the matrices and vectors are
divided (virtually) into num_shards shards.
|
int |
getNumThreads()
The number of threads to use.
|
com.google.protobuf.Parser<PrimalDualHybridGradientParams> |
getParserForType() |
PrimalDualHybridGradientParams.PresolveOptions |
getPresolveOptions()
optional .operations_research.pdlp.PrimalDualHybridGradientParams.PresolveOptions presolve_options = 16; |
PrimalDualHybridGradientParams.PresolveOptionsOrBuilder |
getPresolveOptionsOrBuilder()
optional .operations_research.pdlp.PrimalDualHybridGradientParams.PresolveOptions presolve_options = 16; |
double |
getPrimalWeightUpdateSmoothing()
This parameter controls exponential smoothing of log(primal_weight) when a
primal weight update occurs (i.e., when the ratio of primal and dual step
sizes is adjusted).
|
int |
getRandomProjectionSeeds(int index)
Seeds for generating (pseudo-)random projections of iterates during
termination checks.
|
int |
getRandomProjectionSeedsCount()
Seeds for generating (pseudo-)random projections of iterates during
termination checks.
|
java.util.List<java.lang.Integer> |
getRandomProjectionSeedsList()
Seeds for generating (pseudo-)random projections of iterates during
termination checks.
|
boolean |
getRecordIterationStats()
If true, the iteration_stats field of the SolveLog output will be populated
at every iteration.
|
PrimalDualHybridGradientParams.RestartStrategy |
getRestartStrategy()
NO_RESTARTS and EVERY_MAJOR_ITERATION occasionally outperform the default.
|
int |
getSerializedSize() |
double |
getSufficientReductionForRestart()
For ADAPTIVE_HEURISTIC and ADAPTIVE_DISTANCE_BASED only: A relative
reduction in the potential function by this amount always triggers a
restart.
|
int |
getTerminationCheckFrequency()
The frequency (based on a counter reset every major iteration) to check for
termination (involves extra work) and log iteration stats.
|
TerminationCriteria |
getTerminationCriteria()
optional .operations_research.pdlp.TerminationCriteria termination_criteria = 1; |
TerminationCriteriaOrBuilder |
getTerminationCriteriaOrBuilder()
optional .operations_research.pdlp.TerminationCriteria termination_criteria = 1; |
boolean |
getUseDiagonalQpTrustRegionSolver()
When solving QPs with diagonal objective matrices, this option can be
turned on to enable an experimental solver that avoids linearization of the
quadratic term.
|
boolean |
getUseFeasibilityPolishing()
If true, periodically runs feasibility polishing, which attempts to move
from latest average iterate to one that is closer to feasibility (i.e., has
smaller primal and dual residuals) while probably increasing the objective
gap.
|
int |
getVerbosityLevel()
The verbosity of logging.
|
boolean |
hasAdaptiveLinesearchParameters()
optional .operations_research.pdlp.AdaptiveLinesearchParams adaptive_linesearch_parameters = 18; |
boolean |
hasDiagonalQpTrustRegionSolverTolerance()
The solve tolerance of the experimental trust region solver for diagonal
QPs, controlling the accuracy of binary search over a one-dimensional
scaling parameter.
|
boolean |
hasHandleSomePrimalGradientsOnFiniteBoundsAsResiduals()
See
https://developers.google.com/optimization/lp/pdlp_math#treating_some_variable_bounds_as_infinite
for a description of this flag.
|
int |
hashCode() |
boolean |
hasInfiniteConstraintBoundThreshold()
Constraint bounds with absolute value at least this threshold are replaced
with infinities.
|
boolean |
hasInitialPrimalWeight()
The initial value of the primal weight (i.e., the ratio of primal and dual
step sizes).
|
boolean |
hasInitialStepSizeScaling()
Scaling factor applied to the initial step size (all step sizes if
linesearch_rule == CONSTANT_STEP_SIZE_RULE).
|
boolean |
hasL2NormRescaling()
If true, applies L_2 norm rescaling after the Ruiz rescaling.
|
boolean |
hasLinesearchRule()
Linesearch rule applied at each major iteration.
|
boolean |
hasLInfRuizIterations()
Number of L_infinity Ruiz rescaling iterations to apply to the constraint
matrix.
|
boolean |
hasLogIntervalSeconds()
Time between iteration-level statistics logging (if `verbosity_level > 1`).
|
boolean |
hasMajorIterationFrequency()
The frequency at which extra work is performed to make major algorithmic
decisions, e.g., performing restarts and updating the primal weight.
|
boolean |
hasMalitskyPockParameters()
optional .operations_research.pdlp.MalitskyPockParams malitsky_pock_parameters = 19; |
boolean |
hasNecessaryReductionForRestart()
For ADAPTIVE_HEURISTIC only: A relative reduction in the potential function
by this amount triggers a restart if, additionally, the quality of the
iterates appears to be getting worse.
|
boolean |
hasNumShards()
For more efficient parallel computation, the matrices and vectors are
divided (virtually) into num_shards shards.
|
boolean |
hasNumThreads()
The number of threads to use.
|
boolean |
hasPresolveOptions()
optional .operations_research.pdlp.PrimalDualHybridGradientParams.PresolveOptions presolve_options = 16; |
boolean |
hasPrimalWeightUpdateSmoothing()
This parameter controls exponential smoothing of log(primal_weight) when a
primal weight update occurs (i.e., when the ratio of primal and dual step
sizes is adjusted).
|
boolean |
hasRecordIterationStats()
If true, the iteration_stats field of the SolveLog output will be populated
at every iteration.
|
boolean |
hasRestartStrategy()
NO_RESTARTS and EVERY_MAJOR_ITERATION occasionally outperform the default.
|
boolean |
hasSufficientReductionForRestart()
For ADAPTIVE_HEURISTIC and ADAPTIVE_DISTANCE_BASED only: A relative
reduction in the potential function by this amount always triggers a
restart.
|
boolean |
hasTerminationCheckFrequency()
The frequency (based on a counter reset every major iteration) to check for
termination (involves extra work) and log iteration stats.
|
boolean |
hasTerminationCriteria()
optional .operations_research.pdlp.TerminationCriteria termination_criteria = 1; |
boolean |
hasUseDiagonalQpTrustRegionSolver()
When solving QPs with diagonal objective matrices, this option can be
turned on to enable an experimental solver that avoids linearization of the
quadratic term.
|
boolean |
hasUseFeasibilityPolishing()
If true, periodically runs feasibility polishing, which attempts to move
from latest average iterate to one that is closer to feasibility (i.e., has
smaller primal and dual residuals) while probably increasing the objective
gap.
|
boolean |
hasVerbosityLevel()
The verbosity of logging.
|
protected com.google.protobuf.GeneratedMessage.FieldAccessorTable |
internalGetFieldAccessorTable() |
boolean |
isInitialized() |
static PrimalDualHybridGradientParams.Builder |
newBuilder() |
static PrimalDualHybridGradientParams.Builder |
newBuilder(PrimalDualHybridGradientParams prototype) |
PrimalDualHybridGradientParams.Builder |
newBuilderForType() |
protected PrimalDualHybridGradientParams.Builder |
newBuilderForType(com.google.protobuf.AbstractMessage.BuilderParent parent) |
static PrimalDualHybridGradientParams |
parseDelimitedFrom(java.io.InputStream input) |
static PrimalDualHybridGradientParams |
parseDelimitedFrom(java.io.InputStream input,
com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
static PrimalDualHybridGradientParams |
parseFrom(byte[] data) |
static PrimalDualHybridGradientParams |
parseFrom(byte[] data,
com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
static PrimalDualHybridGradientParams |
parseFrom(java.nio.ByteBuffer data) |
static PrimalDualHybridGradientParams |
parseFrom(java.nio.ByteBuffer data,
com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
static PrimalDualHybridGradientParams |
parseFrom(com.google.protobuf.ByteString data) |
static PrimalDualHybridGradientParams |
parseFrom(com.google.protobuf.ByteString data,
com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
static PrimalDualHybridGradientParams |
parseFrom(com.google.protobuf.CodedInputStream input) |
static PrimalDualHybridGradientParams |
parseFrom(com.google.protobuf.CodedInputStream input,
com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
static PrimalDualHybridGradientParams |
parseFrom(java.io.InputStream input) |
static PrimalDualHybridGradientParams |
parseFrom(java.io.InputStream input,
com.google.protobuf.ExtensionRegistryLite extensionRegistry) |
static com.google.protobuf.Parser<PrimalDualHybridGradientParams> |
parser() |
PrimalDualHybridGradientParams.Builder |
toBuilder() |
void |
writeTo(com.google.protobuf.CodedOutputStream output) |
canUseUnsafe, computeStringSize, computeStringSizeNoTag, emptyBooleanList, emptyDoubleList, emptyFloatList, emptyIntList, emptyList, emptyLongList, getAllFields, getDescriptorForType, getField, getOneofFieldDescriptor, getRepeatedField, getRepeatedFieldCount, getUnknownFields, hasField, hasOneof, internalGetMapField, internalGetMapFieldReflection, isStringEmpty, makeMutableCopy, makeMutableCopy, mergeFromAndMakeImmutableInternal, newFileScopedGeneratedExtension, newInstance, newMessageScopedGeneratedExtension, parseDelimitedWithIOException, parseDelimitedWithIOException, parseUnknownField, parseUnknownFieldProto3, parseWithIOException, parseWithIOException, parseWithIOException, parseWithIOException, serializeBooleanMapTo, serializeIntegerMapTo, serializeLongMapTo, serializeStringMapTo, writeReplace, writeString, writeStringNoTagfindInitializationErrors, getInitializationErrorString, hashFields, toStringaddAll, checkByteStringIsUtf8, toByteArray, toByteString, writeDelimitedTo, writeToclone, finalize, getClass, notify, notifyAll, wait, wait, waitpublic static final int TERMINATION_CRITERIA_FIELD_NUMBER
public static final int NUM_THREADS_FIELD_NUMBER
public static final int NUM_SHARDS_FIELD_NUMBER
public static final int RECORD_ITERATION_STATS_FIELD_NUMBER
public static final int VERBOSITY_LEVEL_FIELD_NUMBER
public static final int LOG_INTERVAL_SECONDS_FIELD_NUMBER
public static final int MAJOR_ITERATION_FREQUENCY_FIELD_NUMBER
public static final int TERMINATION_CHECK_FREQUENCY_FIELD_NUMBER
public static final int RESTART_STRATEGY_FIELD_NUMBER
public static final int PRIMAL_WEIGHT_UPDATE_SMOOTHING_FIELD_NUMBER
public static final int INITIAL_PRIMAL_WEIGHT_FIELD_NUMBER
public static final int PRESOLVE_OPTIONS_FIELD_NUMBER
public static final int L_INF_RUIZ_ITERATIONS_FIELD_NUMBER
public static final int L2_NORM_RESCALING_FIELD_NUMBER
public static final int SUFFICIENT_REDUCTION_FOR_RESTART_FIELD_NUMBER
public static final int NECESSARY_REDUCTION_FOR_RESTART_FIELD_NUMBER
public static final int LINESEARCH_RULE_FIELD_NUMBER
public static final int ADAPTIVE_LINESEARCH_PARAMETERS_FIELD_NUMBER
public static final int MALITSKY_POCK_PARAMETERS_FIELD_NUMBER
public static final int INITIAL_STEP_SIZE_SCALING_FIELD_NUMBER
public static final int RANDOM_PROJECTION_SEEDS_FIELD_NUMBER
public static final int INFINITE_CONSTRAINT_BOUND_THRESHOLD_FIELD_NUMBER
public static final int HANDLE_SOME_PRIMAL_GRADIENTS_ON_FINITE_BOUNDS_AS_RESIDUALS_FIELD_NUMBER
public static final int USE_DIAGONAL_QP_TRUST_REGION_SOLVER_FIELD_NUMBER
public static final int DIAGONAL_QP_TRUST_REGION_SOLVER_TOLERANCE_FIELD_NUMBER
public static final int USE_FEASIBILITY_POLISHING_FIELD_NUMBER
public static final com.google.protobuf.Descriptors.Descriptor getDescriptor()
protected com.google.protobuf.GeneratedMessage.FieldAccessorTable internalGetFieldAccessorTable()
internalGetFieldAccessorTable in class com.google.protobuf.GeneratedMessagepublic boolean hasTerminationCriteria()
optional .operations_research.pdlp.TerminationCriteria termination_criteria = 1;hasTerminationCriteria in interface PrimalDualHybridGradientParamsOrBuilderpublic TerminationCriteria getTerminationCriteria()
optional .operations_research.pdlp.TerminationCriteria termination_criteria = 1;getTerminationCriteria in interface PrimalDualHybridGradientParamsOrBuilderpublic TerminationCriteriaOrBuilder getTerminationCriteriaOrBuilder()
optional .operations_research.pdlp.TerminationCriteria termination_criteria = 1;getTerminationCriteriaOrBuilder in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasNumThreads()
The number of threads to use. Must be positive. Try various values of num_threads, up to the number of physical cores. Performance may not be monotonically increasing with the number of threads because of memory bandwidth limitations.
optional int32 num_threads = 2 [default = 1];hasNumThreads in interface PrimalDualHybridGradientParamsOrBuilderpublic int getNumThreads()
The number of threads to use. Must be positive. Try various values of num_threads, up to the number of physical cores. Performance may not be monotonically increasing with the number of threads because of memory bandwidth limitations.
optional int32 num_threads = 2 [default = 1];getNumThreads in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasNumShards()
For more efficient parallel computation, the matrices and vectors are divided (virtually) into num_shards shards. Results are computed independently for each shard and then combined. As a consequence, the order of computation, and hence floating point roundoff, depends on the number of shards so reproducible results require using the same value for num_shards. However, for efficiency num_shards should a be at least num_threads, and preferably at least 4*num_threads to allow better load balancing. If num_shards is positive, the computation will use that many shards. Otherwise a default that depends on num_threads will be used.
optional int32 num_shards = 27 [default = 0];hasNumShards in interface PrimalDualHybridGradientParamsOrBuilderpublic int getNumShards()
For more efficient parallel computation, the matrices and vectors are divided (virtually) into num_shards shards. Results are computed independently for each shard and then combined. As a consequence, the order of computation, and hence floating point roundoff, depends on the number of shards so reproducible results require using the same value for num_shards. However, for efficiency num_shards should a be at least num_threads, and preferably at least 4*num_threads to allow better load balancing. If num_shards is positive, the computation will use that many shards. Otherwise a default that depends on num_threads will be used.
optional int32 num_shards = 27 [default = 0];getNumShards in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasRecordIterationStats()
If true, the iteration_stats field of the SolveLog output will be populated at every iteration. Note that we only compute solution statistics at termination checks. Setting this parameter to true may substantially increase the size of the output.
optional bool record_iteration_stats = 3;hasRecordIterationStats in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean getRecordIterationStats()
If true, the iteration_stats field of the SolveLog output will be populated at every iteration. Note that we only compute solution statistics at termination checks. Setting this parameter to true may substantially increase the size of the output.
optional bool record_iteration_stats = 3;getRecordIterationStats in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasVerbosityLevel()
The verbosity of logging. 0: No informational logging. (Errors are logged.) 1: Summary statistics only. No iteration-level details. 2: A table of iteration-level statistics is logged. (See ToShortString() in primal_dual_hybrid_gradient.cc). 3: A more detailed table of iteration-level statistics is logged. (See ToString() in primal_dual_hybrid_gradient.cc). 4: For iteration-level details, prints the statistics of both the average (prefixed with A) and the current iterate (prefixed with C). Also prints internal algorithmic state and details. Logging at levels 2-4 also includes messages from level 1.
optional int32 verbosity_level = 26 [default = 0];hasVerbosityLevel in interface PrimalDualHybridGradientParamsOrBuilderpublic int getVerbosityLevel()
The verbosity of logging. 0: No informational logging. (Errors are logged.) 1: Summary statistics only. No iteration-level details. 2: A table of iteration-level statistics is logged. (See ToShortString() in primal_dual_hybrid_gradient.cc). 3: A more detailed table of iteration-level statistics is logged. (See ToString() in primal_dual_hybrid_gradient.cc). 4: For iteration-level details, prints the statistics of both the average (prefixed with A) and the current iterate (prefixed with C). Also prints internal algorithmic state and details. Logging at levels 2-4 also includes messages from level 1.
optional int32 verbosity_level = 26 [default = 0];getVerbosityLevel in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasLogIntervalSeconds()
Time between iteration-level statistics logging (if `verbosity_level > 1`). Since iteration-level statistics are only generated when performing termination checks, logs will be generated from next termination check after `log_interval_seconds` have elapsed. Should be >= 0.0. 0.0 (the default) means log statistics at every termination check.
optional double log_interval_seconds = 31 [default = 0];hasLogIntervalSeconds in interface PrimalDualHybridGradientParamsOrBuilderpublic double getLogIntervalSeconds()
Time between iteration-level statistics logging (if `verbosity_level > 1`). Since iteration-level statistics are only generated when performing termination checks, logs will be generated from next termination check after `log_interval_seconds` have elapsed. Should be >= 0.0. 0.0 (the default) means log statistics at every termination check.
optional double log_interval_seconds = 31 [default = 0];getLogIntervalSeconds in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasMajorIterationFrequency()
The frequency at which extra work is performed to make major algorithmic decisions, e.g., performing restarts and updating the primal weight. Major iterations also trigger a termination check. For best performance using the NO_RESTARTS or EVERY_MAJOR_ITERATION rule, one should perform a log-scale grid search over this parameter, for example, over powers of two. ADAPTIVE_HEURISTIC is mostly insensitive to this value.
optional int32 major_iteration_frequency = 4 [default = 64];hasMajorIterationFrequency in interface PrimalDualHybridGradientParamsOrBuilderpublic int getMajorIterationFrequency()
The frequency at which extra work is performed to make major algorithmic decisions, e.g., performing restarts and updating the primal weight. Major iterations also trigger a termination check. For best performance using the NO_RESTARTS or EVERY_MAJOR_ITERATION rule, one should perform a log-scale grid search over this parameter, for example, over powers of two. ADAPTIVE_HEURISTIC is mostly insensitive to this value.
optional int32 major_iteration_frequency = 4 [default = 64];getMajorIterationFrequency in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasTerminationCheckFrequency()
The frequency (based on a counter reset every major iteration) to check for termination (involves extra work) and log iteration stats. Termination checks do not affect algorithmic progress unless termination is triggered.
optional int32 termination_check_frequency = 5 [default = 64];hasTerminationCheckFrequency in interface PrimalDualHybridGradientParamsOrBuilderpublic int getTerminationCheckFrequency()
The frequency (based on a counter reset every major iteration) to check for termination (involves extra work) and log iteration stats. Termination checks do not affect algorithmic progress unless termination is triggered.
optional int32 termination_check_frequency = 5 [default = 64];getTerminationCheckFrequency in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasRestartStrategy()
NO_RESTARTS and EVERY_MAJOR_ITERATION occasionally outperform the default. If using a strategy other than ADAPTIVE_HEURISTIC, you must also tune major_iteration_frequency.
optional .operations_research.pdlp.PrimalDualHybridGradientParams.RestartStrategy restart_strategy = 6 [default = ADAPTIVE_HEURISTIC];hasRestartStrategy in interface PrimalDualHybridGradientParamsOrBuilderpublic PrimalDualHybridGradientParams.RestartStrategy getRestartStrategy()
NO_RESTARTS and EVERY_MAJOR_ITERATION occasionally outperform the default. If using a strategy other than ADAPTIVE_HEURISTIC, you must also tune major_iteration_frequency.
optional .operations_research.pdlp.PrimalDualHybridGradientParams.RestartStrategy restart_strategy = 6 [default = ADAPTIVE_HEURISTIC];getRestartStrategy in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasPrimalWeightUpdateSmoothing()
This parameter controls exponential smoothing of log(primal_weight) when a primal weight update occurs (i.e., when the ratio of primal and dual step sizes is adjusted). At 0.0, the primal weight will be frozen at its initial value and there will be no dynamic updates in the algorithm. At 1.0, there is no smoothing in the updates. The default of 0.5 generally performs well, but has been observed on occasion to trigger unstable swings in the primal weight. We recommend also trying 0.0 (disabling primal weight updates), in which case you must also tune initial_primal_weight.
optional double primal_weight_update_smoothing = 7 [default = 0.5];hasPrimalWeightUpdateSmoothing in interface PrimalDualHybridGradientParamsOrBuilderpublic double getPrimalWeightUpdateSmoothing()
This parameter controls exponential smoothing of log(primal_weight) when a primal weight update occurs (i.e., when the ratio of primal and dual step sizes is adjusted). At 0.0, the primal weight will be frozen at its initial value and there will be no dynamic updates in the algorithm. At 1.0, there is no smoothing in the updates. The default of 0.5 generally performs well, but has been observed on occasion to trigger unstable swings in the primal weight. We recommend also trying 0.0 (disabling primal weight updates), in which case you must also tune initial_primal_weight.
optional double primal_weight_update_smoothing = 7 [default = 0.5];getPrimalWeightUpdateSmoothing in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasInitialPrimalWeight()
The initial value of the primal weight (i.e., the ratio of primal and dual
step sizes). The primal weight remains fixed throughout the solve if
primal_weight_update_smoothing = 0.0. If unset, the default is the ratio of
the norm of the objective vector to the L2 norm of the combined constraint
bounds vector (as defined above). If this ratio is not finite and positive,
then the default is 1.0 instead. For tuning, try powers of 10, for example,
from 10^{-6} to 10^6.
optional double initial_primal_weight = 8;hasInitialPrimalWeight in interface PrimalDualHybridGradientParamsOrBuilderpublic double getInitialPrimalWeight()
The initial value of the primal weight (i.e., the ratio of primal and dual
step sizes). The primal weight remains fixed throughout the solve if
primal_weight_update_smoothing = 0.0. If unset, the default is the ratio of
the norm of the objective vector to the L2 norm of the combined constraint
bounds vector (as defined above). If this ratio is not finite and positive,
then the default is 1.0 instead. For tuning, try powers of 10, for example,
from 10^{-6} to 10^6.
optional double initial_primal_weight = 8;getInitialPrimalWeight in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasPresolveOptions()
optional .operations_research.pdlp.PrimalDualHybridGradientParams.PresolveOptions presolve_options = 16;hasPresolveOptions in interface PrimalDualHybridGradientParamsOrBuilderpublic PrimalDualHybridGradientParams.PresolveOptions getPresolveOptions()
optional .operations_research.pdlp.PrimalDualHybridGradientParams.PresolveOptions presolve_options = 16;getPresolveOptions in interface PrimalDualHybridGradientParamsOrBuilderpublic PrimalDualHybridGradientParams.PresolveOptionsOrBuilder getPresolveOptionsOrBuilder()
optional .operations_research.pdlp.PrimalDualHybridGradientParams.PresolveOptions presolve_options = 16;getPresolveOptionsOrBuilder in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasLInfRuizIterations()
Number of L_infinity Ruiz rescaling iterations to apply to the constraint matrix. Zero disables this rescaling pass. Recommended values to try when tuning are 0, 5, and 10.
optional int32 l_inf_ruiz_iterations = 9 [default = 5];hasLInfRuizIterations in interface PrimalDualHybridGradientParamsOrBuilderpublic int getLInfRuizIterations()
Number of L_infinity Ruiz rescaling iterations to apply to the constraint matrix. Zero disables this rescaling pass. Recommended values to try when tuning are 0, 5, and 10.
optional int32 l_inf_ruiz_iterations = 9 [default = 5];getLInfRuizIterations in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasL2NormRescaling()
If true, applies L_2 norm rescaling after the Ruiz rescaling. Heuristically this has been found to help convergence.
optional bool l2_norm_rescaling = 10 [default = true];hasL2NormRescaling in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean getL2NormRescaling()
If true, applies L_2 norm rescaling after the Ruiz rescaling. Heuristically this has been found to help convergence.
optional bool l2_norm_rescaling = 10 [default = true];getL2NormRescaling in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasSufficientReductionForRestart()
For ADAPTIVE_HEURISTIC and ADAPTIVE_DISTANCE_BASED only: A relative reduction in the potential function by this amount always triggers a restart. Must be between 0.0 and 1.0.
optional double sufficient_reduction_for_restart = 11 [default = 0.1];hasSufficientReductionForRestart in interface PrimalDualHybridGradientParamsOrBuilderpublic double getSufficientReductionForRestart()
For ADAPTIVE_HEURISTIC and ADAPTIVE_DISTANCE_BASED only: A relative reduction in the potential function by this amount always triggers a restart. Must be between 0.0 and 1.0.
optional double sufficient_reduction_for_restart = 11 [default = 0.1];getSufficientReductionForRestart in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasNecessaryReductionForRestart()
For ADAPTIVE_HEURISTIC only: A relative reduction in the potential function by this amount triggers a restart if, additionally, the quality of the iterates appears to be getting worse. The value must be in the interval [sufficient_reduction_for_restart, 1). Smaller values make restarts less frequent, and larger values make them more frequent.
optional double necessary_reduction_for_restart = 17 [default = 0.9];hasNecessaryReductionForRestart in interface PrimalDualHybridGradientParamsOrBuilderpublic double getNecessaryReductionForRestart()
For ADAPTIVE_HEURISTIC only: A relative reduction in the potential function by this amount triggers a restart if, additionally, the quality of the iterates appears to be getting worse. The value must be in the interval [sufficient_reduction_for_restart, 1). Smaller values make restarts less frequent, and larger values make them more frequent.
optional double necessary_reduction_for_restart = 17 [default = 0.9];getNecessaryReductionForRestart in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasLinesearchRule()
Linesearch rule applied at each major iteration.
optional .operations_research.pdlp.PrimalDualHybridGradientParams.LinesearchRule linesearch_rule = 12 [default = ADAPTIVE_LINESEARCH_RULE];hasLinesearchRule in interface PrimalDualHybridGradientParamsOrBuilderpublic PrimalDualHybridGradientParams.LinesearchRule getLinesearchRule()
Linesearch rule applied at each major iteration.
optional .operations_research.pdlp.PrimalDualHybridGradientParams.LinesearchRule linesearch_rule = 12 [default = ADAPTIVE_LINESEARCH_RULE];getLinesearchRule in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasAdaptiveLinesearchParameters()
optional .operations_research.pdlp.AdaptiveLinesearchParams adaptive_linesearch_parameters = 18;hasAdaptiveLinesearchParameters in interface PrimalDualHybridGradientParamsOrBuilderpublic AdaptiveLinesearchParams getAdaptiveLinesearchParameters()
optional .operations_research.pdlp.AdaptiveLinesearchParams adaptive_linesearch_parameters = 18;getAdaptiveLinesearchParameters in interface PrimalDualHybridGradientParamsOrBuilderpublic AdaptiveLinesearchParamsOrBuilder getAdaptiveLinesearchParametersOrBuilder()
optional .operations_research.pdlp.AdaptiveLinesearchParams adaptive_linesearch_parameters = 18;getAdaptiveLinesearchParametersOrBuilder in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasMalitskyPockParameters()
optional .operations_research.pdlp.MalitskyPockParams malitsky_pock_parameters = 19;hasMalitskyPockParameters in interface PrimalDualHybridGradientParamsOrBuilderpublic MalitskyPockParams getMalitskyPockParameters()
optional .operations_research.pdlp.MalitskyPockParams malitsky_pock_parameters = 19;getMalitskyPockParameters in interface PrimalDualHybridGradientParamsOrBuilderpublic MalitskyPockParamsOrBuilder getMalitskyPockParametersOrBuilder()
optional .operations_research.pdlp.MalitskyPockParams malitsky_pock_parameters = 19;getMalitskyPockParametersOrBuilder in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasInitialStepSizeScaling()
Scaling factor applied to the initial step size (all step sizes if linesearch_rule == CONSTANT_STEP_SIZE_RULE).
optional double initial_step_size_scaling = 25 [default = 1];hasInitialStepSizeScaling in interface PrimalDualHybridGradientParamsOrBuilderpublic double getInitialStepSizeScaling()
Scaling factor applied to the initial step size (all step sizes if linesearch_rule == CONSTANT_STEP_SIZE_RULE).
optional double initial_step_size_scaling = 25 [default = 1];getInitialStepSizeScaling in interface PrimalDualHybridGradientParamsOrBuilderpublic java.util.List<java.lang.Integer> getRandomProjectionSeedsList()
Seeds for generating (pseudo-)random projections of iterates during termination checks. For each seed, the projection of the primal and dual solutions onto random planes in primal and dual space will be computed and added the IterationStats if record_iteration_stats is true. The random planes generated will be determined by the seeds, the primal and dual dimensions, and num_threads.
repeated int32 random_projection_seeds = 28 [packed = true];getRandomProjectionSeedsList in interface PrimalDualHybridGradientParamsOrBuilderpublic int getRandomProjectionSeedsCount()
Seeds for generating (pseudo-)random projections of iterates during termination checks. For each seed, the projection of the primal and dual solutions onto random planes in primal and dual space will be computed and added the IterationStats if record_iteration_stats is true. The random planes generated will be determined by the seeds, the primal and dual dimensions, and num_threads.
repeated int32 random_projection_seeds = 28 [packed = true];getRandomProjectionSeedsCount in interface PrimalDualHybridGradientParamsOrBuilderpublic int getRandomProjectionSeeds(int index)
Seeds for generating (pseudo-)random projections of iterates during termination checks. For each seed, the projection of the primal and dual solutions onto random planes in primal and dual space will be computed and added the IterationStats if record_iteration_stats is true. The random planes generated will be determined by the seeds, the primal and dual dimensions, and num_threads.
repeated int32 random_projection_seeds = 28 [packed = true];getRandomProjectionSeeds in interface PrimalDualHybridGradientParamsOrBuilderindex - The index of the element to return.public boolean hasInfiniteConstraintBoundThreshold()
Constraint bounds with absolute value at least this threshold are replaced with infinities. NOTE: This primarily affects the relative convergence criteria. A smaller value makes the relative convergence criteria stronger. It also affects the problem statistics LOG()ed at the start of the run, and the default initial primal weight, since that is based on the norm of the bounds.
optional double infinite_constraint_bound_threshold = 22 [default = inf];hasInfiniteConstraintBoundThreshold in interface PrimalDualHybridGradientParamsOrBuilderpublic double getInfiniteConstraintBoundThreshold()
Constraint bounds with absolute value at least this threshold are replaced with infinities. NOTE: This primarily affects the relative convergence criteria. A smaller value makes the relative convergence criteria stronger. It also affects the problem statistics LOG()ed at the start of the run, and the default initial primal weight, since that is based on the norm of the bounds.
optional double infinite_constraint_bound_threshold = 22 [default = inf];getInfiniteConstraintBoundThreshold in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasHandleSomePrimalGradientsOnFiniteBoundsAsResiduals()
See https://developers.google.com/optimization/lp/pdlp_math#treating_some_variable_bounds_as_infinite for a description of this flag.
optional bool handle_some_primal_gradients_on_finite_bounds_as_residuals = 29 [default = true];hasHandleSomePrimalGradientsOnFiniteBoundsAsResiduals in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean getHandleSomePrimalGradientsOnFiniteBoundsAsResiduals()
See https://developers.google.com/optimization/lp/pdlp_math#treating_some_variable_bounds_as_infinite for a description of this flag.
optional bool handle_some_primal_gradients_on_finite_bounds_as_residuals = 29 [default = true];getHandleSomePrimalGradientsOnFiniteBoundsAsResiduals in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasUseDiagonalQpTrustRegionSolver()
When solving QPs with diagonal objective matrices, this option can be turned on to enable an experimental solver that avoids linearization of the quadratic term. The `diagonal_qp_solver_accuracy` parameter controls the solve accuracy. TODO(user): Turn this option on by default for quadratic programs after numerical evaluation.
optional bool use_diagonal_qp_trust_region_solver = 23 [default = false];hasUseDiagonalQpTrustRegionSolver in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean getUseDiagonalQpTrustRegionSolver()
When solving QPs with diagonal objective matrices, this option can be turned on to enable an experimental solver that avoids linearization of the quadratic term. The `diagonal_qp_solver_accuracy` parameter controls the solve accuracy. TODO(user): Turn this option on by default for quadratic programs after numerical evaluation.
optional bool use_diagonal_qp_trust_region_solver = 23 [default = false];getUseDiagonalQpTrustRegionSolver in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasDiagonalQpTrustRegionSolverTolerance()
The solve tolerance of the experimental trust region solver for diagonal QPs, controlling the accuracy of binary search over a one-dimensional scaling parameter. Smaller values imply smaller relative error of the final solution vector. TODO(user): Find an expression for the final relative error.
optional double diagonal_qp_trust_region_solver_tolerance = 24 [default = 1e-08];hasDiagonalQpTrustRegionSolverTolerance in interface PrimalDualHybridGradientParamsOrBuilderpublic double getDiagonalQpTrustRegionSolverTolerance()
The solve tolerance of the experimental trust region solver for diagonal QPs, controlling the accuracy of binary search over a one-dimensional scaling parameter. Smaller values imply smaller relative error of the final solution vector. TODO(user): Find an expression for the final relative error.
optional double diagonal_qp_trust_region_solver_tolerance = 24 [default = 1e-08];getDiagonalQpTrustRegionSolverTolerance in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean hasUseFeasibilityPolishing()
If true, periodically runs feasibility polishing, which attempts to move from latest average iterate to one that is closer to feasibility (i.e., has smaller primal and dual residuals) while probably increasing the objective gap. This is useful primarily when the feasibility tolerances are fairly tight and the objective gap tolerance is somewhat looser. Note that this does not change the termination criteria, but rather can help achieve the termination criteria more quickly when the objective gap is not as important as feasibility. `use_feasibility_polishing` cannot be used with glop presolve, and requires `handle_some_primal_gradients_on_finite_bounds_as_residuals == false`. `use_feasibility_polishing` can only be used with linear programs. Feasibility polishing runs two separate phases, primal feasibility and dual feasibility. The primal feasibility phase runs PDHG on the primal feasibility problem (obtained by changing the objective vector to all zeros), using the average primal iterate and zero dual (which is optimal for the primal feasibility problem) as the initial solution. The dual feasibility phase runs PDHG on the dual feasibility problem (obtained by changing all finite variable and constraint bounds to zero), using the average dual iterate and zero primal (which is optimal for the dual feasibility problem) as the initial solution. The primal solution from the primal feasibility phase and dual solution from the dual feasibility phase are then combined (forming a solution of type `POINT_TYPE_FEASIBILITY_POLISHING_SOLUTION`) and checked against the termination criteria.
optional bool use_feasibility_polishing = 30 [default = false];hasUseFeasibilityPolishing in interface PrimalDualHybridGradientParamsOrBuilderpublic boolean getUseFeasibilityPolishing()
If true, periodically runs feasibility polishing, which attempts to move from latest average iterate to one that is closer to feasibility (i.e., has smaller primal and dual residuals) while probably increasing the objective gap. This is useful primarily when the feasibility tolerances are fairly tight and the objective gap tolerance is somewhat looser. Note that this does not change the termination criteria, but rather can help achieve the termination criteria more quickly when the objective gap is not as important as feasibility. `use_feasibility_polishing` cannot be used with glop presolve, and requires `handle_some_primal_gradients_on_finite_bounds_as_residuals == false`. `use_feasibility_polishing` can only be used with linear programs. Feasibility polishing runs two separate phases, primal feasibility and dual feasibility. The primal feasibility phase runs PDHG on the primal feasibility problem (obtained by changing the objective vector to all zeros), using the average primal iterate and zero dual (which is optimal for the primal feasibility problem) as the initial solution. The dual feasibility phase runs PDHG on the dual feasibility problem (obtained by changing all finite variable and constraint bounds to zero), using the average dual iterate and zero primal (which is optimal for the dual feasibility problem) as the initial solution. The primal solution from the primal feasibility phase and dual solution from the dual feasibility phase are then combined (forming a solution of type `POINT_TYPE_FEASIBILITY_POLISHING_SOLUTION`) and checked against the termination criteria.
optional bool use_feasibility_polishing = 30 [default = false];getUseFeasibilityPolishing in interface PrimalDualHybridGradientParamsOrBuilderpublic final boolean isInitialized()
isInitialized in interface com.google.protobuf.MessageLiteOrBuilderisInitialized in class com.google.protobuf.GeneratedMessagepublic void writeTo(com.google.protobuf.CodedOutputStream output)
throws java.io.IOException
writeTo in interface com.google.protobuf.MessageLitewriteTo in class com.google.protobuf.GeneratedMessagejava.io.IOExceptionpublic int getSerializedSize()
getSerializedSize in interface com.google.protobuf.MessageLitegetSerializedSize in class com.google.protobuf.GeneratedMessagepublic boolean equals(java.lang.Object obj)
equals in interface com.google.protobuf.Messageequals in class com.google.protobuf.AbstractMessagepublic int hashCode()
hashCode in interface com.google.protobuf.MessagehashCode in class com.google.protobuf.AbstractMessagepublic static PrimalDualHybridGradientParams parseFrom(java.nio.ByteBuffer data) throws com.google.protobuf.InvalidProtocolBufferException
com.google.protobuf.InvalidProtocolBufferExceptionpublic static PrimalDualHybridGradientParams parseFrom(java.nio.ByteBuffer data, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws com.google.protobuf.InvalidProtocolBufferException
com.google.protobuf.InvalidProtocolBufferExceptionpublic static PrimalDualHybridGradientParams parseFrom(com.google.protobuf.ByteString data) throws com.google.protobuf.InvalidProtocolBufferException
com.google.protobuf.InvalidProtocolBufferExceptionpublic static PrimalDualHybridGradientParams parseFrom(com.google.protobuf.ByteString data, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws com.google.protobuf.InvalidProtocolBufferException
com.google.protobuf.InvalidProtocolBufferExceptionpublic static PrimalDualHybridGradientParams parseFrom(byte[] data) throws com.google.protobuf.InvalidProtocolBufferException
com.google.protobuf.InvalidProtocolBufferExceptionpublic static PrimalDualHybridGradientParams parseFrom(byte[] data, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws com.google.protobuf.InvalidProtocolBufferException
com.google.protobuf.InvalidProtocolBufferExceptionpublic static PrimalDualHybridGradientParams parseFrom(java.io.InputStream input) throws java.io.IOException
java.io.IOExceptionpublic static PrimalDualHybridGradientParams parseFrom(java.io.InputStream input, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws java.io.IOException
java.io.IOExceptionpublic static PrimalDualHybridGradientParams parseDelimitedFrom(java.io.InputStream input) throws java.io.IOException
java.io.IOExceptionpublic static PrimalDualHybridGradientParams parseDelimitedFrom(java.io.InputStream input, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws java.io.IOException
java.io.IOExceptionpublic static PrimalDualHybridGradientParams parseFrom(com.google.protobuf.CodedInputStream input) throws java.io.IOException
java.io.IOExceptionpublic static PrimalDualHybridGradientParams parseFrom(com.google.protobuf.CodedInputStream input, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws java.io.IOException
java.io.IOExceptionpublic PrimalDualHybridGradientParams.Builder newBuilderForType()
newBuilderForType in interface com.google.protobuf.MessagenewBuilderForType in interface com.google.protobuf.MessageLitepublic static PrimalDualHybridGradientParams.Builder newBuilder()
public static PrimalDualHybridGradientParams.Builder newBuilder(PrimalDualHybridGradientParams prototype)
public PrimalDualHybridGradientParams.Builder toBuilder()
toBuilder in interface com.google.protobuf.MessagetoBuilder in interface com.google.protobuf.MessageLiteprotected PrimalDualHybridGradientParams.Builder newBuilderForType(com.google.protobuf.AbstractMessage.BuilderParent parent)
newBuilderForType in class com.google.protobuf.AbstractMessagepublic static PrimalDualHybridGradientParams getDefaultInstance()
public static com.google.protobuf.Parser<PrimalDualHybridGradientParams> parser()
public com.google.protobuf.Parser<PrimalDualHybridGradientParams> getParserForType()
getParserForType in interface com.google.protobuf.MessagegetParserForType in interface com.google.protobuf.MessageLitegetParserForType in class com.google.protobuf.GeneratedMessagepublic PrimalDualHybridGradientParams getDefaultInstanceForType()
getDefaultInstanceForType in interface com.google.protobuf.MessageLiteOrBuildergetDefaultInstanceForType in interface com.google.protobuf.MessageOrBuilderCopyright © 2024. All rights reserved.