hide long namesshow long names
hide short namesshow short names
Integer type:  int32  int64  nag_int  show int32  show int32  show int64  show int64  show nag_int  show nag_int

PDF version (NAG web site, 64-bit version, 64-bit version)
Chapter Contents
Chapter Introduction
NAG Toolbox

NAG Toolbox: nag_specfun_airy_ai_real_vector (s17au)

Purpose

nag_specfun_airy_ai_real_vector (s17au) returns an array of values for the Airy function, Ai(x)Ai(x).

Syntax

[f, ivalid, ifail] = s17au(x, 'n', n)
[f, ivalid, ifail] = nag_specfun_airy_ai_real_vector(x, 'n', n)

Description

nag_specfun_airy_ai_real_vector (s17au) evaluates an approximation to the Airy function, Ai(xi)Ai(xi) for an array of arguments xixi, for i = 1,2,,ni=1,2,,n. It is based on a number of Chebyshev expansions:
For x < 5x<-5,
Ai(x) = (a(t)sinzb(t)cosz)/((x)1 / 4)
Ai(x)=a(t)sinz-b(t)cosz(-x)1/4
where z = π/4 + (2/3)sqrt(x3)z= π4+ 23-x3, and a(t)a(t) and b(t)b(t) are expansions in the variable t = 2(5/x)31t=-2 ( 5x) 3-1.
For 5x0-5x0,
Ai(x) = f(t)xg(t),
Ai(x)=f(t)-xg(t),
where ff and gg are expansions in t = 2(x/5)31.t=-2 ( x5) 3-1.
For 0 < x < 4.50<x<4.5,
Ai(x) = e3x / 2y(t),
Ai(x)=e-3x/2y(t),
where yy is an expansion in t = 4x / 91t=4x/9-1.
For 4.5x < 94.5x<9,
Ai(x) = e5x / 2u(t),
Ai(x)=e-5x/2u(t),
where uu is an expansion in t = 4x / 93t=4x/9-3.
For x9x9,
Ai(x) = (ezv(t))/(x1 / 4),
Ai(x)=e-zv(t)x1/4,
where z = (2/3)sqrt(x3)z= 23x3 and vv is an expansion in t = 2 (18/z)1t=2 ( 18z)-1.
For |x| < machine precision|x|<machine precision, the result is set directly to Ai(0)Ai(0). This both saves time and guards against underflow in intermediate calculations.
For large negative arguments, it becomes impossible to calculate the phase of the oscillatory function with any precision and so the function must fail. This occurs if x < (3/(2ε))2 / 3x<- ( 32ε ) 2/3 , where εε is the machine precision.
For large positive arguments, where AiAi decays in an essentially exponential manner, there is a danger of underflow so the function must fail.

References

Abramowitz M and Stegun I A (1972) Handbook of Mathematical Functions (3rd Edition) Dover Publications

Parameters

Compulsory Input Parameters

1:     x(n) – double array
n, the dimension of the array, must satisfy the constraint n0n0.
The argument xixi of the function, for i = 1,2,,ni=1,2,,n.

Optional Input Parameters

1:     n – int64int32nag_int scalar
Default: The dimension of the array x.
nn, the number of points.
Constraint: n0n0.

Input Parameters Omitted from the MATLAB Interface

None.

Output Parameters

1:     f(n) – double array
Ai(xi)Ai(xi), the function values.
2:     ivalid(n) – int64int32nag_int array
ivalid(i)ivalidi contains the error code for xixi, for i = 1,2,,ni=1,2,,n.
ivalid(i) = 0ivalidi=0
No error.
ivalid(i) = 1ivalidi=1
xixi is too large and positive. f(i)fi contains zero. The threshold value is the same as for ifail = 1ifail=1 in nag_specfun_airy_ai_real (s17ag), as defined in the Users' Note for your implementation.
ivalid(i) = 2ivalidi=2
xixi is too large and negative. f(i)fi contains zero. The threshold value is the same as for ifail = 2ifail=2 in nag_specfun_airy_ai_real (s17ag), as defined in the Users' Note for your implementation.
3:     ifail – int64int32nag_int scalar
ifail = 0ifail=0 unless the function detects an error (see [Error Indicators and Warnings]).

Error Indicators and Warnings

Errors or warnings detected by the function:

Cases prefixed with W are classified as warnings and do not generate an error of type NAG:error_n. See nag_issue_warnings.

W ifail = 1ifail=1
On entry, at least one value of x was invalid.
Check ivalid for more information.
  ifail = 2ifail=2
Constraint: n0n0.

Accuracy

For negative arguments the function is oscillatory and hence absolute error is the appropriate measure. In the positive region the function is essentially exponential-like and here relative error is appropriate. The absolute error, EE, and the relative error, εε, are related in principle to the relative error in the argument, δδ, by
E |xAi(x)|δ, ε |( x Ai(x) )/(Ai(x))|δ.
E | x Ai(x) |δ, ε | x Ai(x) Ai(x) |δ.
In practice, approximate equality is the best that can be expected. When δδ, εε or EE is of the order of the machine precision, the errors in the result will be somewhat larger.
For small xx, errors are strongly damped by the function and hence will be bounded by the machine precision.
For moderate negative xx, the error behaviour is oscillatory but the amplitude of the error grows like
amplitude (E/δ) (|x|5 / 4)/(sqrt(π)).
amplitude (Eδ ) |x|5/4π.
However the phase error will be growing roughly like (2/3)sqrt(|x|3) 23|x|3 and hence all accuracy will be lost for large negative arguments due to the impossibility of calculating sin and cos to any accuracy if (2/3)sqrt(|x|3) > 1/δ 23|x|3> 1δ .
For large positive arguments, the relative error amplification is considerable:
ε/δsqrt(x3).
ε δ x3.
This means a loss of roughly two decimal places accuracy for arguments in the region of 2020. However very large arguments are not possible due to the danger of setting underflow and so the errors are limited in practice.

Further Comments

None.

Example

function nag_specfun_airy_ai_real_vector_example
x = [-10; -1; 0; 1; 5; 10; 20];
[f, ivalid, ifail] = nag_specfun_airy_ai_real_vector(x);
fprintf('\n    X           Y\n');
for i=1:numel(x)
  fprintf('%12.3e%12.3e%5d\n', x(i), f(i), ivalid(i));
end
 

    X           Y
  -1.000e+01   4.024e-02    0
  -1.000e+00   5.356e-01    0
   0.000e+00   3.550e-01    0
   1.000e+00   1.353e-01    0
   5.000e+00   1.083e-04    0
   1.000e+01   1.105e-10    0
   2.000e+01   1.692e-27    0

function s17au_example
x = [-10; -1; 0; 1; 5; 10; 20];
[f, ivalid, ifail] = s17au(x);
fprintf('\n    X           Y\n');
for i=1:numel(x)
  fprintf('%12.3e%12.3e%5d\n', x(i), f(i), ivalid(i));
end
 

    X           Y
  -1.000e+01   4.024e-02    0
  -1.000e+00   5.356e-01    0
   0.000e+00   3.550e-01    0
   1.000e+00   1.353e-01    0
   5.000e+00   1.083e-04    0
   1.000e+01   1.105e-10    0
   2.000e+01   1.692e-27    0


PDF version (NAG web site, 64-bit version, 64-bit version)
Chapter Contents
Chapter Introduction
NAG Toolbox

© The Numerical Algorithms Group Ltd, Oxford, UK. 2009–2013