Modifier and Type | Field and Description |
---|---|
double |
cplexEpGap
Description copied from Cplex: Sets a relative tolerance on the gap between the best integer objective and the objective of the best node remaining.
|
double |
cplexTiLim
Description copied from Cplex: Sets the maximum time, in seconds, for a call to an optimizer.
|
double |
cplexTreLim
Description copied from Cplex: Sets an absolute upper limit on the size (in megabytes, uncompressed) of the branch-and-cut tree.
|
int |
maxCPUSec
The maximum time limit in seconds for the whole program.
|
int |
maxLocalSearchIter
The maximum number of iterations for each local search instance.
|
int |
maxStochIter
The maximum number of iterations to solve the main stochastic problem.
|
int |
numCPU |
boolean |
resetTargetSLCP
While optimizing by cutting-planes, setting this parameter to
true
will change the service level (SL) targets (lowered by targetSLTolCall and
targetSLTolGlobal parameters) to their original SL targets. |
boolean |
solveIP |
double |
targetSLTolCall
The SL target tolerance level for each call type during the cutting-plane
optimization.
|
double |
targetSLTolGlobal
The SL target tolerance level for the aggregate SL.
|
boolean |
useLocalSearch
Set to
true to execute a local search on the
stochastic solution at the end of the program. |
boolean |
verbose
Set to
true to print information to screen. |
Constructor and Description |
---|
StochCPParams() |
public int numCPU
public boolean solveIP
public double cplexTreLim
solveIP
parameter is set to true
.
Set a negative value to use Cplex's default value.
This parameter only affects the main stochastic staffing problem. It
does not affect the instances of CuttingPlaneOnly
.
public double cplexTiLim
solveIP
parameter is set to true
.
Set a negative value to use Cplex's default value.
This parameter only affects the main stochastic staffing problem. It
does not affect the instances of CuttingPlaneOnly
.
public double cplexEpGap
solveIP
parameter is set to true
.
Set a negative value to use Cplex's default value.
This parameter only affects the main stochastic staffing problem. It
does not affect the instances of CuttingPlaneOnly
.
public double targetSLTolCall
This parameter only affects the instances of CuttingPlaneOnly
.
It does not apply to the local search.
public double targetSLTolGlobal
targetSLTolCall
.
We suggest a value between 10% and 20% of the global SL target.
The user must take care when setting this parameter.
The lowered SL target should not be below the parameter
CuttingPlaneParams.pivotPerCallSL
.
Otherwise, the algorithm CuttingPlaneOnly
may loop indefinitely.
public boolean resetTargetSLCP
true
will change the service level (SL) targets (lowered by targetSLTolCall
and
targetSLTolGlobal
parameters) to their original SL targets.
This is corresponds to executing the program CuttingPlaneOnly
independently
with lowered SL targets, then optimize all scenarios simultaneously
with cutting-planes with the original SL targets afterward.
When this parameter is to false
, the cutting-plane method will only
use the lowered SL targets and never the original SL targets.
Local search should to be executed in order to find a feasible solution,
by setting useLocalSearch
to true
.
We recommend to set to true
.
public int maxStochIter
CuttingPlaneOnly
.public int maxCPUSec
CuttingPlaneOnly
and the local search.
Note that executing many parallel threads will not multiply the running time count.public boolean useLocalSearch
true
to execute a local search on the
stochastic solution at the end of the program.
We recommend to set to true
.public int maxLocalSearchIter
public boolean verbose
true
to print information to screen.