Solution
A solution represents the evaluated inputs for on a problem definition. Solution also implements the EstimatedResponseIfc interface by delegating to the supplied estimated objective function. The FeasibilityIfc interface is implemented by delegating to the supplied input map.
Parameters
the inputs (name,value) pairs associated with the solution
the estimated objective function from the simulation oracle
the estimates of the responses associated with the response constraints
the iteration number of the solver request. That is, the number of times that the simulation oracle has been asked to evaluate (any) input.
Constructors
Properties
The estimated (average) value of the objective function
The estimated (average) value of the objective function but rounded to the problem's granularity for the objective function
The value of the penalized objective function but rounded to the problem's granularity for the objective function
Allows comparison of solutions by the estimated objective function
The penalized objective function. That is, the estimated objective function plus the total penalty associated with violating the constraints.
Allows comparison of solutions by the estimated objective function
The current value of the penalty function
The response estimate averages
The total penalty associated with violating the response constraints
The number of times that the response has been sampled. The sample size of the response estimates.
The standard deviations of the estimated responses
The variance of the estimated responses
The violation amount for each response constraint
Functions
Converts the solution to a map of name, value pairs. The returned values include the input map, the estimated objective function, and the estimated responses.
Tests if each response constraint is feasible. If all tests are feasible, then the solution is considered response feasible.
Computes a one-sided upper confidence interval for each response constraint to test if the interval contains zero. If the upper limit of the interval is less than 0.0, then we can be confident that response constraint is feasible. The individual confidence interval upper limits are based on a one-sided confidence interval on the mean response assuming normality. The upper limit is computed as (x_bar - b + t(level, n-1)*s/sqrt(n)) assuming a less-than constraint. The individual confidence interval levels are adjusted to meet the overall level of confidence.
Converts the solution to an instance of a ResponseMap