The NLP Procedure 
Computational Resources 
Since nonlinear optimization is an iterative process that depends on many factors, it is difficult to estimate how much computer time is necessary to compute an optimal solution satisfying one of the termination criteria. The MAXTIME=, MAXITER=, and MAXFUNC= options can be used to restrict the amount of real time, the number of iterations, and the number of function calls in a single run of PROC NLP.
In each iteration , the NRRIDG and LEVMAR techniques use symmetric Householder transformations to decompose the Hessian (crossproduct Jacobian) matrix ,
to compute the (Newton) search direction :
The QUADAS, TRUREG, NEWRAP, and HYQUAN techniques use the Cholesky decomposition to solve the same linear system while computing the search direction. The QUANEW, DBLDOG, CONGRA, and NMSIMP techniques do not need to invert or decompose a Hessian or crossproduct Jacobian matrix and thus require fewer computational resources then the first group of techniques.
The larger the problem, the more time is spent computing function values and derivatives. Therefore, many researchers compare optimization techniques by counting and comparing the respective numbers of function, gradient, and Hessian (crossproduct Jacobian) evaluations. You can save computer time and memory by specifying derivatives (using the GRADIENT, JACOBIAN, CRPJAC, or HESSIAN statement) since you will typically produce a more efficient representation than the internal derivative compiler.
Finitedifference approximations of the derivatives are expensive since they require additional function or gradient calls.
Forwarddifference formulas:
Firstorder derivatives: additional function calls are needed.
Secondorder derivatives based on function calls only: for a dense Hessian, additional function calls are needed.
Secondorder derivatives based on gradient calls: additional gradient calls are needed.
Centraldifference formulas:
Firstorder derivatives: additional function calls are needed.
Secondorder derivatives based on function calls only: for a dense Hessian, additional function calls are needed.
Secondorder derivatives based on gradient: additional gradient calls are needed.
Many applications need considerably more time for computing secondorder derivatives (Hessian matrix) than for firstorder derivatives (gradient). In such cases, a (dual) quasiNewton or conjugate gradient technique is recommended, which does not require secondorder derivatives.
The following table shows for each optimization technique which derivatives are needed (FOD: firstorder derivatives; SOD: secondorder derivatives), what kinds of constraints are supported (BC: boundary constraints; LIC: linear constraints), and the minimal memory (number of double floating point numbers) required. For various reasons, there are additionally about double floating point numbers needed.
Quadratic Programming 
FOD 
SOD 
BC 
LIC 
Memory 
LICOMP 
 
 
x 
x 

QUADAS 
 
 
x 
x 

General Optimization 
FOD 
SOD 
BC 
LIC 
Memory 
TRUREG 
x 
x 
x 
x 

NEWRAP 
x 
x 
x 
x 

NRRIDG 
x 
x 
x 
x 

QUANEW 
x 
 
x 
x 

DBLDOG 
x 
 
x 
x 

CONGRA 
x 
 
x 
x 

NMSIMP 
 
 
x 
x 

Least Squares 
FOD 
SOD 
BC 
LIC 
Memory 
LEVMAR 
x 
 
x 
x 

HYQUAN 
x 
 
x 
x 

Notes:
Here, denotes the number of parameters, the squared number of parameters, and .
The value of is the product of the number of functions specified in the MIN, MAX, or LSQ statement and the maximum number of observations in each BY group of a DATA= input data set. The following table also contains the number of variables in the DATA= data set that are used in the program statements.
For a diagonal Hessian matrix, the term in QUADAS, TRUREG, NEWRAP, and NRRIDG is replaced by .
If the TRUREG, NRRIDG, or NEWRAP method is used to minimize a least squares problem, the second derivatives are replaced by the crossproduct Jacobian matrix.
The memory needed by the TECH=NONE specification depends on the output specifications (typically, it needs double floating point numbers and an additional if the Jacobian matrix is required).
The total amount of memory needed to run an optimization technique consists of the techniquespecific memory listed in the preceding table, plus additional blocks of memory as shown in the following table.
double 
int 
long 
8byte 

Basic Requirement 




DATA= data set 

 
 

JACOBIAN statement 

 
 
 
CRPJAC statement 

 
 
 
HESSIAN statement 

 
 
 
COV= option 

 
 
 
Scaling vector 

 
 
 
BOUNDS statement 


 
 
Bounds in INEST= 

 
 
 
LINCON and TRUREG 


 
 
LINCON and other 


 
 
Notes:
For TECH=LICOMP, the total amount of memory needed for the linear or boundary constrained case is , where is the number of constraints.
The amount of memory needed to specify derivatives with a GRADIENT, JACOBIAN, CRPJAC, or HESSIAN statement (shown in this table) is small compared to that needed for using the internal function compiler to compute the derivatives. This is especially so for secondorder derivatives.
If the CONGRA technique is used, specifying the GRADCHECK=DETAIL option requires an additional double floating point numbers to store the finitedifference Hessian matrix.
Copyright © SAS Institute, Inc. All Rights Reserved.