Integer type:  int32  int64  nag_int  show int32  show int32  show int64  show int64  show nag_int  show nag_int

Chapter Contents
Chapter Introduction
NAG Toolbox

## Purpose

nag_quad_md_sphere (d01fd) calculates an approximation to a definite integral in up to 30$30$ dimensions, using the method of Sag and Szekeres (see Sag and Szekeres (1964)). The region of integration is an n$n$-sphere, or by built-in transformation via the unit n$n$-cube, any product region.

## Syntax

[result, ncalls, ifail] = d01fd(ndim, f, sigma, region, limit, 'r0', r0, 'u', u)
[result, ncalls, ifail] = nag_quad_md_sphere(ndim, f, sigma, region, limit, 'r0', r0, 'u', u)

## Description

nag_quad_md_sphere (d01fd) calculates an approximation to
 ∫ f(x1,x2, … ,xn)dx1dx2 ⋯ dxn n-sphere of radius  σ
(1)
or, more generally,
 d1 dn ∫ dx1 ⋯ ∫ dxnf(x1, … ,xn) c1 cn
$∫ c1 d1 dx1 ⋯ ∫ cn dn dxn f (x1,…,xn)$
(2)
where each ci${c}_{i}$ and di${d}_{i}$ may be functions of xj${x}_{j}$ (j < i)$\left(j.
The function uses the method of Sag and Szekeres (1964), which exploits a property of the shifted p$p$-point trapezoidal rule, namely, that it integrates exactly all polynomials of degree < p$\text{} (see Krylov (1962)). An attempt is made to induce periodicity in the integrand by making a parameterised transformation to the unit n$n$-sphere. The Jacobian of the transformation and all its direct derivatives vanish rapidly towards the surface of the unit n$n$-sphere, so that, except for functions which have strong singularities on the boundary, the resulting integrand will be pseudo-periodic. In addition, the variation in the integrand can be considerably reduced, causing the trapezoidal rule to perform well.
Integrals of the form (1) are transformed to the unit n$n$-sphere by the change of variables:
 xi = yi σ/r tanh((ur)/(1 − r2)) $xi = yi σr tanh( ur 1-r2 )$
where r2 = i = 1nyi2${r}^{2}=\sum _{i=1}^{n}{y}_{i}^{2}$ and u$u$ is an adjustable parameter.
Integrals of the form (2) are first of all transformed to the n$n$-cube [1,1]n${\left[-1,1\right]}^{n}$ by a linear change of variables
 xi = ((di + ci) + (di − ci)yi) / 2 $xi=((di+ci)+(di-ci)yi)/2$
and then to the unit sphere by a further change of variables
 yi = tanh((uzi)/(1 − r)) $yi=tanh(uzi 1-r )$
where r2 = i = 1nzi2${r}^{2}=\sum _{i=1}^{n}{z}_{i}^{2}$ and u$u$ is again an adjustable parameter.
The parameter u$u$ in these transformations determines how the transformed integrand is distributed between the origin and the surface of the unit n$n$-sphere. A typical value of u$u$ is 1.5$1.5$. For larger u$u$, the integrand is concentrated toward the centre of the unit n$n$-sphere, while for smaller u$u$ it is concentrated toward the perimeter.
In performing the integration over the unit n$n$-sphere by the trapezoidal rule, a displaced equidistant grid of size h$h$ is constructed. The points of the mesh lie on concentric layers of radius
 ri = h/4sqrt(n + 8(i − 1)),  i = 1,2,3, … . $ri=h4n+8(i-1), i=1,2,3,….$
The function requires you to specify an approximate maximum number of points to be used, and then computes the largest number of whole layers to be used, subject to an upper limit of 400$400$ layers.
In practice, the rapidly-decreasing Jacobian makes it unnecessary to include the whole unit n$n$-sphere and the integration region is limited by a user-specified cut-off radius r0 < 1${r}_{0}<1$. The grid-spacing h$h$ is determined by r0${r}_{0}$ and the number of layers to be used. A typical value of r0${r}_{0}$ is 0.8$0.8$.
Some experimentation may be required with the choice of r0${r}_{0}$ (which determines how much of the unit n$n$-sphere is included) and u$u$ (which determines how the transformed integrand is distributed between the origin and surface of the unit n$n$-sphere), to obtain best results for particular families of integrals. This matter is discussed further in Section [Further Comments].

## References

Krylov V I (1962) Approximate Calculation of Integrals (trans A H Stroud) Macmillan
Sag T W and Szekeres G (1964) Numerical evaluation of high-dimensional integrals Math. Comput. 18 245–253

## Parameters

### Compulsory Input Parameters

1:     ndim – int64int32nag_int scalar
n$n$, the number of dimensions of the integral.
Constraint: 1ndim30$1\le {\mathbf{ndim}}\le 30$.
2:     f – function handle or string containing name of m-file
f must return the value of the integrand f$f$ at a given point.
[result] = f(ndim, x)

Input Parameters

1:     ndim – int64int32nag_int scalar
n$n$, the number of dimensions of the integral.
2:     x(ndim) – double array
The coordinates of the point at which the integrand f$f$ must be evaluated.

Output Parameters

1:     result – double scalar
The result of the function.
3:     sigma – double scalar
Indicates the region of integration.
sigma0.0${\mathbf{sigma}}\ge 0.0$
The integration is carried out over the n$n$-sphere of radius sigma, centred at the origin.
sigma < 0.0${\mathbf{sigma}}<0.0$
The integration is carried out over the product region described by region.
4:     region – function handle or string containing name of m-file
If sigma < 0.0${\mathbf{sigma}}<0.0$, region must evaluate the limits of integration in any dimension.
[c, d] = region(ndim, x, j)

Input Parameters

1:     ndim – int64int32nag_int scalar
n$n$, the number of dimensions of the integral.
2:     x(ndim) – double array
x(1),,x(j1)${\mathbf{x}}\left(1\right),\dots ,{\mathbf{x}}\left(j-1\right)$ contain the current values of the first (j1)$\left(j-1\right)$ variables, which may be used if necessary in calculating cj${c}_{j}$ and dj${d}_{j}$.
3:     j – int64int32nag_int scalar
The index j$j$ for which the limits of the range of integration are required.

Output Parameters

1:     c – double scalar
The lower limit cj${c}_{j}$ of the range of xj${x}_{j}$.
2:     d – double scalar
The upper limit dj${d}_{j}$ of the range of xj${x}_{j}$.
If sigma0.0${\mathbf{sigma}}\ge 0.0$, region is not called by nag_quad_md_sphere (d01fd), string 'd01fdv'
5:     limit – int64int32nag_int scalar
The approximate maximum number of integrand evaluations to be used.
Constraint: limit100${\mathbf{limit}}\ge 100$.

### Optional Input Parameters

1:     r0 – double scalar
The cut-off radius on the unit n$n$-sphere, which may be regarded as an adjustable parameter of the method.
Default: 0.8$0.8$
Constraint: 0.0 < r0 < 1.0$0.0<{\mathbf{r0}}<1.0$.
2:     u – double scalar
Must specify an adjustable parameter of the transformation to the unit n$n$-sphere.
Default: 1.5$1.5$
Constraint: u > 0.0${\mathbf{u}}>0.0$.

None.

### Output Parameters

1:     result – double scalar
The approximation to the integral I$I$.
2:     ncalls – int64int32nag_int scalar
3:     ifail – int64int32nag_int scalar
${\mathrm{ifail}}={\mathbf{0}}$ unless the function detects an error (see [Error Indicators and Warnings]).

## Error Indicators and Warnings

Errors or warnings detected by the function:
ifail = 1${\mathbf{ifail}}=1$
 On entry, ndim < 1${\mathbf{ndim}}<1$, or ndim > 30${\mathbf{ndim}}>30$.
ifail = 2${\mathbf{ifail}}=2$
 On entry, limit < 100${\mathbf{limit}}<100$.
ifail = 3${\mathbf{ifail}}=3$
 On entry, r0 ≤ 0.0${\mathbf{r0}}\le 0.0$, or r0 ≥ 1.0${\mathbf{r0}}\ge 1.0$.
ifail = 4${\mathbf{ifail}}=4$
 On entry, u ≤ 0.0${\mathbf{u}}\le 0.0$.

## Accuracy

No error estimate is returned, but results may be verified by repeating with an increased value of limit (provided that this causes an increase in the returned value of ncalls).

The time taken by nag_quad_md_sphere (d01fd) will be approximately proportional to the returned value of ncalls, which, except in the circumstances outlined in (b) below, will be close to the given value of limit.
(a) Choice of r0${r}_{0}$ and u$u$
If the chosen combination of r0${r}_{0}$ and u$u$ is too large in relation to the machine accuracy it is possible that some of the points generated in the original region of integration may transform into points in the unit n$n$-sphere which lie too close to the boundary surface to be distinguished from it to machine accuracy (despite the fact that r0 < 1${r}_{0}<1$). To be specific, the combination of r0${r}_{0}$ and u$u$ is too large if
 (ur0)/(1 − r02) > 0.3465(t − 1),   if ​sigma ≥ 0.0, $ur0 1-r02 >0.3465(t-1), if ​sigma≥0.0,$
or
 (ur0)/(1 − r0) > 0.3465(t − 1),   if ​ sigma < 0.0, $ur0 1-r0 > 0.3465(t- 1), if ​ sigma< 0.0,$
where t$t$ is the number of bits in the mantissa of a double number.
The contribution of such points to the integral is neglected. This may be justified by appeal to the fact that the Jacobian of the transformation rapidly approaches zero towards the surface. Neglect of these points avoids the occurrence of overflow with integrands which are infinite on the boundary.
(b) Values of limit and ncalls
limit is an approximate upper limit to the number of integrand evaluations, and may not be chosen less than 100$100$. There are two circumstances when the returned value of ncalls (the actual number of evaluations used) may be significantly less than limit.
Firstly, as explained in (a), an unsuitably large combination of r0${r}_{0}$ and u$u$ may result in some of the points being unusable. Such points are not included in the returned value of ncalls.
Secondly, no more than 400$400$ layers will ever be used, no matter how high limit is set. This places an effective upper limit on ncalls as follows:
 n = 1 : 56 n = 2 : 1252 n = 3 : 23690 n = 4 : 394528 n = 5 : 5956906
$n=1: 56 n=2: 1252 n=3: 23690 n=4: 394528 n=5: 5956906$

## Example

```function nag_quad_md_sphere_example
ndim = int64(3);
sigma = 1.5;
limit = int64(8000);
[result, ncalls, ifail] = ...
nag_quad_md_sphere(ndim, @f, sigma, @region, limit, 'r0', 0.9)

function result = f(ndim,x)
result = 2.25;
for i = 1:double(ndim)
result = result - x(i)^2;
end
result=1/sqrt(abs(result));
function [c,d] = region(ndim, x, j)
c = -1.5;
d = 1.5;
if (j > 1)
sm = 2.25;
for i = 1:double(j-1)
sm = sm - x(i)*x(i);
end
d = sqrt(abs(sm));
c = -d;
end
```
```

result =

22.1679

ncalls =

8026

ifail =

0

```
```function d01fd_example
ndim = int64(3);
sigma = 1.5;
limit = int64(8000);
[result, ncalls, ifail] = ...
d01fd(ndim, @f, sigma, @region, limit, 'r0', 0.9)

function result = f(ndim,x)
result = 2.25;
for i = 1:double(ndim)
result = result - x(i)^2;
end
result=1/sqrt(abs(result));
function [c,d] = region(ndim, x, j)
c = -1.5;
d = 1.5;
if (j > 1)
sm = 2.25;
for i = 1:double(j-1)
sm = sm - x(i)*x(i);
end
d = sqrt(abs(sm));
c = -d;
end
```
```

result =

22.1679

ncalls =

8026

ifail =

0

```