Iterative Display
Introduction
The iterative display is a table of statistics describing the calculations in each iteration of a solver. The statistics depend on both the solver and the solver algorithm. The table appears in the MATLAB®Command Window when you run solvers with appropriate options. For more information about iterations, seeIterations and Function Counts.
Obtain the iterative display by usingoptimoptions
with theDisplay
option set to'iter'
or'iter-detailed'
. For example:
options = optimoptions(@fminunc,'Display','iter','Algorithm','quasi-newton'); [x fval exitflag output] = fminunc(@sin,0,options);
First-order Iteration Func-count f(x) Step-size optimality 0 2 0 1 1 4 -0.841471 1 0.54 2 8 -1 0.484797 0.000993 3 10 -1 1 5.62e-05 4 12 -1 1 0 Local minimum found. Optimization completed because the size of the gradient is less than the value of the optimality tolerance.
The iterative display is available for all solvers except:
lsqlin
'trust-region-reflective'
algorithmlsqnonneg
quadprog
'trust-region-reflective'
algorithm
Common Headings
This table lists some common headings of iterative display.
Heading | Information Displayed |
---|---|
|
Current objective function value; for |
|
First-order optimality measure (seeFirst-Order Optimality Measure) |
|
Number of function evaluations; seeIterations and Function Counts |
|
Iteration number; seeIterations and Function Counts |
|
Size of the current step (size is the Euclidean norm, or 2-norm). For the |
Function-Specific Headings
The tables in this section describe headings of the iterative display whose meaning is specific to the optimization function you are using.
fgoalattain, fmincon, fminimax, and fseminf
This table describes the headings specific tofgoalattain
,fmincon
,fminimax
, andfseminf
.
fgoalattain, fmincon, fminimax, or fseminf Heading | Information Displayed |
---|---|
|
Value of the attainment factor for |
|
Number of conjugate gradient iterations taken in the current iteration (seePreconditioned Conjugate Gradient Method) |
|
Gradient of the objective function along the search direction |
|
Maximum constraint violation, where satisfied inequality constraints count as |
|
Multiplicative factor that scales the search direction (seeEquation 29) |
|
Maximum violation among all constraints, both internally constructed and user-provided; can be negative when no constraint is binding |
|
Objective function value of the nonlinear programming reformulation of the minimax problem for |
|
Hessian update procedures:
For more information, seeUpdating the Hessian Matrix. QP subproblem procedures:
|
|
Multiplicative factor that scales the search direction (seeEquation 29) |
|
Current trust-region radius |
fminbnd and fzero
This table describes the headings specific tofminbnd
andfzero
.
fminbnd or fzero Heading | Information Displayed |
---|---|
|
Procedures for
Procedures for
|
|
当前点的算法 |
fminsearch
This table describes the headings specific tofminsearch
.
fminsearch Heading | Information Displayed |
---|---|
|
Minimum function value in the current simplex |
|
Simplex procedure at the current iteration. Procedures include:
For details, seefminsearch Algorithm. |
fminunc
This table describes the headings specific tofminunc
.
fminunc Heading | Information Displayed |
---|---|
|
Number of conjugate gradient iterations taken in the current iteration (seePreconditioned Conjugate Gradient Method) |
|
Multiplicative factor that scales the search direction (seeEquation 11) |
Thefminunc
'quasi-newton'
algorithm can issue askipped update
message to the right of theFirst-order optimality
column. This message means thatfminunc
did not update its Hessian estimate, because the resulting matrix would not have been positive definite. The message usually indicates that the objective function is not smooth at the current point.
fsolve
This table describes the headings specific tofsolve
.
fsolve Heading | Information Displayed |
---|---|
|
Gradient of the function along the search direction |
|
λkvalue defined inLevenberg-Marquardt Method |
|
Residual (sum of squares) of the function |
|
Current trust-region radius (change in the norm of the trust-region radius) |
intlinprog
This table describes the headings specific tointlinprog
.
intlinprog Heading | Information Displayed |
---|---|
|
Cumulative number of explored nodes |
|
Time in seconds since |
|
Number of integer feasible points found |
|
Objective function value of the best integer feasible point found. This value is an upper bound for the final objective function value |
|
where
Note Although you specify |
linprog
This table describes the headings specific tolinprog
. Each algorithm has its own iterative display.
linprog Heading | Information Displayed |
---|---|
|
Primal infeasibility, a measure of the constraint violations, which should be zero at a solution. For definitions, seePredictor-Corrector( |
|
Dual infeasibility, a measure of the derivative of the Lagrangian, which should be zero at a solution. For the definition of the Lagrangian, seePredictor-Corrector. For the definition of dual infeasibility, seePredictor-Corrector( |
|
Upper bound feasibility.{x}means thosexwith finite upper bounds. This value is theruresidual inInterior-Point-Legacy Linear Programming. |
|
Duality gap (seeInterior-Point-Legacy Linear Programming) between the primal objective and the dual objective. |
|
Total relative error, described at the end ofMain Algorithm |
|
A measure of the Lagrange multipliers times distance from the bounds, which should be zero at a solution. See thercvariable inStopping Conditions. |
|
Time in seconds that |
lsqlin
Thelsqlin
'interior-point'
iterative display is inherited from thequadprog
iterative display. The relationship between these functions is explained inLinear Least Squares: Interior-Point or Active-Set. For iterative display details, seequadprog.
lsqnonlin and lsqcurvefit
This table describes the headings specific tolsqnonlin
andlsqcurvefit
.
lsqnonlin or lsqcurvefit Heading | Information Displayed |
---|---|
|
Gradient of the function along the search direction |
|
λkvalue defined inLevenberg-Marquardt Method |
|
Value of the squared 2-norm of the residual at |
|
Residual vector of the function |
quadprog
This table describes the headings specific toquadprog
. Only the'interior-point-convex'
algorithm has the iterative display.
quadprog Heading | Information Displayed |
---|---|
|
Primal infeasibility, defined as |
|
Dual infeasibility, defined as |
|
的最大绝对值的测量滞后range multipliers of inactive inequalities, which should be zero at a solution. This quantity isginInfeasibility Detection. |