# NAG Library Routine Document

## 1Purpose

f04mff updates the solution of the equations $Tx=b$, where $T$ is a real symmetric positive definite Toeplitz matrix.

## 2Specification

Fortran Interface
 Subroutine f04mff ( n, t, b, x, p, work,
 Integer, Intent (In) :: n Integer, Intent (Inout) :: ifail Real (Kind=nag_wp), Intent (In) :: t(0:*), b(*) Real (Kind=nag_wp), Intent (Inout) :: x(*), work(*) Real (Kind=nag_wp), Intent (Out) :: p
#include nagmk26.h
 void f04mff_ (const Integer *n, const double t[], const double b[], double x[], double *p, double work[], Integer *ifail)

## 3Description

f04mff solves the equations
 $Tnxn=bn,$
where ${T}_{n}$ is the $n$ by $n$ symmetric positive definite Toeplitz matrix
 $Tn= τ0 τ1 τ2 … τn-1 τ1 τ0 τ1 … τn-2 τ2 τ1 τ0 … τn-3 . . . . τn-1 τn-2 τn-3 … τ0$
and ${b}_{n}$ is the $n$-element vector ${b}_{n}={\left({\beta }_{1}{\beta }_{2}\dots {\beta }_{n}\right)}^{\mathrm{T}}$, given the solution of the equations
 $Tn-1xn-1=bn-1.$
This routine will normally be used to successively solve the equations
 $Tkxk=bk, k= 1,2,…,n.$
If it is desired to solve the equations for a single value of $n$, then routine f04fff may be called. This routine uses the method of Levinson (see Levinson (1947) and Golub and Van Loan (1996)).
Bunch J R (1985) Stability of methods for solving Toeplitz systems of equations SIAM J. Sci. Statist. Comput. 6 349–364
Bunch J R (1987) The weak and strong stability of algorithms in numerical linear algebra Linear Algebra Appl. 88/89 49–66
Cybenko G (1980) The numerical stability of the Levinson–Durbin algorithm for Toeplitz systems of equations SIAM J. Sci. Statist. Comput. 1 303–319
Golub G H and Van Loan C F (1996) Matrix Computations (3rd Edition) Johns Hopkins University Press, Baltimore
Levinson N (1947) The Weiner RMS error criterion in filter design and prediction J. Math. Phys. 25 261–278

## 5Arguments

1:     $\mathbf{n}$ – IntegerInput
On entry: the order of the Toeplitz matrix $T$.
Constraint: ${\mathbf{n}}\ge 0$. When ${\mathbf{n}}=0$, an immediate return is effected.
2:     $\mathbf{t}\left(0:*\right)$ – Real (Kind=nag_wp) arrayInput
Note: the dimension of the array t must be at least $\mathrm{max}\phantom{\rule{0.125em}{0ex}}\left(1,{\mathbf{n}}\right)$.
On entry: ${\mathbf{t}}\left(\mathit{i}\right)$ must contain the value ${\tau }_{\mathit{i}}$, for $\mathit{i}=0,1,\dots ,{\mathbf{n}}-1$.
Constraint: ${\mathbf{t}}\left(0\right)>0.0$. Note that if this is not true, the Toeplitz matrix cannot be positive definite.
3:     $\mathbf{b}\left(*\right)$ – Real (Kind=nag_wp) arrayInput
Note: the dimension of the array b must be at least $\mathrm{max}\phantom{\rule{0.125em}{0ex}}\left(1,{\mathbf{n}}\right)$.
On entry: the right-hand side vector ${b}_{n}$.
4:     $\mathbf{x}\left(*\right)$ – Real (Kind=nag_wp) arrayInput/Output
Note: the dimension of the array x must be at least $\mathrm{max}\phantom{\rule{0.125em}{0ex}}\left(1,{\mathbf{n}}\right)$.
On entry: with ${\mathbf{n}}>1$ the ($n-1$) elements of the solution vector ${x}_{n-1}$ as returned by a previous call to f04mff. The element ${\mathbf{x}}\left({\mathbf{n}}\right)$ need not be specified.
On exit: the solution vector ${x}_{n}$.
5:     $\mathbf{p}$ – Real (Kind=nag_wp)Output
On exit: the reflection coefficient ${p}_{n-1}$. (See Section 9.)
6:     $\mathbf{work}\left(*\right)$ – Real (Kind=nag_wp) arrayInput/Output
Note: the dimension of the array work must be at least $\mathrm{max}\phantom{\rule{0.125em}{0ex}}\left(1,2×{\mathbf{n}}-1\right)$.
On entry: with ${\mathbf{n}}>2$ the elements of work should be as returned from a previous call to f04mff with (${\mathbf{n}}-1$) as the argument n.
On exit: the first (${\mathbf{n}}-1$) elements of work contain the solution to the Yule–Walker equations
 $Tn-1yn-1=-tn-1,$
where ${t}_{n-1}={\left({\tau }_{1}{\tau }_{2}\dots {\tau }_{n-1}\right)}^{{\mathbf{t}}}$.
7:     $\mathbf{ifail}$ – IntegerInput/Output
On entry: ifail must be set to $0$, $-1\text{​ or ​}1$. If you are unfamiliar with this argument you should refer to Section 3.4 in How to Use the NAG Library and its Documentation for details.
For environments where it might be inappropriate to halt program execution when an error is detected, the value $-1\text{​ or ​}1$ is recommended. If the output of error messages is undesirable, then the value $1$ is recommended. Otherwise, if you are not familiar with this argument, the recommended value is $0$. When the value $-\mathbf{1}\text{​ or ​}\mathbf{1}$ is used it is essential to test the value of ifail on exit.
On exit: ${\mathbf{ifail}}={\mathbf{0}}$ unless the routine detects an error or a warning has been flagged (see Section 6).

## 6Error Indicators and Warnings

If on entry ${\mathbf{ifail}}=0$ or $-1$, explanatory error messages are output on the current error message unit (as defined by x04aaf).
Errors or warnings detected by the routine:
${\mathbf{ifail}}=-1$
 On entry, ${\mathbf{n}}<0$, or ${\mathbf{t}}\left(0\right)\le 0.0$.
${\mathbf{ifail}}=1$
The Toeplitz matrix ${T}_{n}$ is not positive definite to working accuracy. If, on exit, p is close to unity, ${T}_{n}$ was probably close to being singular.
${\mathbf{ifail}}=-99$
See Section 3.9 in How to Use the NAG Library and its Documentation for further information.
${\mathbf{ifail}}=-399$
Your licence key may have expired or may not have been installed correctly.
See Section 3.8 in How to Use the NAG Library and its Documentation for further information.
${\mathbf{ifail}}=-999$
Dynamic memory allocation failed.
See Section 3.7 in How to Use the NAG Library and its Documentation for further information.

## 7Accuracy

The computed solution of the equations certainly satisfies
 $r=Tnxn-bn,$
where ${‖r‖}_{1}$ is approximately bounded by
 $r1≤cεCTn,$
$c$ being a modest function of $n$, $\epsilon$ being the machine precision and $C\left(T\right)$ being the condition number of $T$ with respect to inversion. This bound is almost certainly pessimistic, but it seems unlikely that the method of Levinson is backward stable, so caution should be exercised when ${T}_{n}$ is ill-conditioned. The following bound on ${T}_{n}^{-1}$ holds:
 $max1∏i=1 n-11-pi2 , 1∏i=1 n-11-pi ≤Tn-11≤∏i=1 n-1 1+pi 1-pi .$
(See Golub and Van Loan (1996).) The norm of ${T}_{n}^{-1}$ may also be estimated using routine f04ydf. For further information on stability issues see Bunch (1985), Bunch (1987), Cybenko (1980) and Golub and Van Loan (1996).

## 8Parallelism and Performance

f04mff makes calls to BLAS and/or LAPACK routines, which may be threaded within the vendor library used by this implementation. Consult the documentation for the vendor library for further information.
Please consult the X06 Chapter Introduction for information on how to control and interrogate the OpenMP environment used within this routine. Please also consult the Users' Note for your implementation for any additional implementation-specific information.

The number of floating-point operations used by this routine is approximately $8n$.
If ${y}_{i}$ is the solution of the equations
 $Tiyi=-τ1τ2…τiT,$
then the reflection coefficient ${p}_{i}$ is defined as the $i$th element of ${y}_{i}$.

## 10Example

This example finds the solution of the equations ${T}_{k}{x}_{k}={b}_{k}$, $k=1,2,3,4$, where
 $T4= 4 3 2 1 3 4 3 2 2 3 4 3 1 2 3 4 and b4= 1 1 1 1 .$

### 10.1Program Text

Program Text (f04mffe.f90)

### 10.2Program Data

Program Data (f04mffe.d)

### 10.3Program Results

Program Results (f04mffe.r)

© The Numerical Algorithms Group Ltd, Oxford, UK. 2017