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
|
\name{solnp}
\alias{solnp}
\title{
Nonlinear optimization using augmented Lagrange method.
}
\description{
The solnp function is based on the solver by Yinyu Ye which solves the general
nonlinear programming problem:\cr
\deqn{\min f(x)}{min f(x)}
\deqn{\mathrm{s.t.}}{s.t.}
\deqn{g(x) = 0}{g(x) = 0}
\deqn{l_h \leq h(x) \leq u_h}{l[h] <= h(x) <= u[h]}
\deqn{l_x \leq x \leq u_x}{l[x] <= x <= u[x]}
where, \eqn{f(x)}, \eqn{g(x)} and \eqn{h(x)} are smooth functions.
}
\usage{
solnp(pars, fun, eqfun = NULL, eqB = NULL, ineqfun = NULL, ineqLB = NULL,
ineqUB = NULL, LB = NULL, UB = NULL, control = list(), ...)
}
\arguments{
\item{pars}{
The starting parameter vector.
}
\item{fun}{
The main function which takes as first argument the parameter vector and returns
a single value.
}
\item{eqfun}{
(Optional) The equality constraint function returning the vector of evaluated
equality constraints.
}
\item{eqB}{
(Optional) The equality constraints.
}
\item{ineqfun}{
(Optional) The inequality constraint function returning the vector of evaluated
inequality constraints.
}
\item{ineqLB}{
(Optional) The lower bound of the inequality constraints.
}
\item{ineqUB}{
(Optional) The upper bound of the inequality constraints.
}
\item{LB}{
(Optional) The lower bound on the parameters.
}
\item{UB}{
(Optional) The upper bound on the parameters.
}
\item{control}{
(Optional) The control list of optimization parameters. See below for details.
}
\item{\dots}{
(Optional) Additional parameters passed to the main, equality or inequality
functions. Note that the main and constraint functions must take the exact same
arguments, irrespective of whether they are used by all of them.
}
}
\details{
The solver belongs to the class of indirect solvers and implements the augmented
Lagrange multiplier method with an SQP interior algorithm.
}
\value{
A list containing the following values:
\item{pars}{Optimal Parameters.}
\item{convergence }{Indicates whether the solver has converged (0) or not
(1 or 2).}
\item{values}{Vector of function values during optimization with last one the
value at the optimal.}
\item{lagrange}{The vector of Lagrange multipliers.}
\item{hessian}{The Hessian of the augmented problem at the optimal solution.}
\item{ineqx0}{The estimated optimal inequality vector of slack variables used
for transforming the inequality into an equality constraint.}
\item{nfuneval}{The number of function evaluations.}
\item{elapsed}{Time taken to compute solution.}
}
\section{Control}{
\describe{
\item{rho}{This is used as a penalty weighting scaler for infeasibility in the
augmented objective function. The higher its value the more the weighting to
bring the solution into the feasible region (default 1). However, very high
values might lead to numerical ill conditioning or significantly slow down
convergence.}
\item{outer.iter}{Maximum number of major (outer) iterations (default 400).}
\item{inner.iter}{Maximum number of minor (inner) iterations (default 800).}
\item{delta}{Relative step size in forward difference evaluation
(default 1.0e-7).}
\item{tol}{ Relative tolerance on feasibility and optimality (default 1e-8).}
\item{trace}{The value of the objective function and the parameters is printed
at every major iteration (default 1).}
}}
\references{
Y.Ye, \emph{Interior algorithms for linear, quadratic, and linearly constrained
non linear programming}, PhD Thesis, Department of EES Stanford University,
Stanford CA.
}
\author{
Alexios Ghalanos and Stefan Theussl\cr
Y.Ye (original matlab version of solnp)
}
\note{
The control parameters \code{tol} and \code{delta} are key in getting any
possibility of successful convergence, therefore it is suggested that the user
change these appropriately to reflect their problem specification.\cr
The solver is a local solver, therefore for problems with rough surfaces and
many local minima there is absolutely no reason to expect anything other than a
local solution.
}
\examples{
# From the original paper by Y.Ye
# see the unit tests for more....
#---------------------------------------------------------------------------------
# POWELL Problem
fn1=function(x)
{
exp(x[1]*x[2]*x[3]*x[4]*x[5])
}
eqn1=function(x){
z1=x[1]*x[1]+x[2]*x[2]+x[3]*x[3]+x[4]*x[4]+x[5]*x[5]
z2=x[2]*x[3]-5*x[4]*x[5]
z3=x[1]*x[1]*x[1]+x[2]*x[2]*x[2]
return(c(z1,z2,z3))
}
x0 = c(-2, 2, 2, -1, -1)
powell=solnp(x0, fun = fn1, eqfun = eqn1, eqB = c(10, 0, -1))
}
\keyword{optimize}
|
