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

# NAG Toolbox: nag_specfun_airy_ai_real (s17ag)

## Purpose

nag_specfun_airy_ai_real (s17ag) returns a value for the Airy function, $\mathrm{Ai}\left(x\right)$, via the function name.

## Syntax

[result, ifail] = s17ag(x)
[result, ifail] = nag_specfun_airy_ai_real(x)

## Description

nag_specfun_airy_ai_real (s17ag) evaluates an approximation to the Airy function, $\mathrm{Ai}\left(x\right)$. It is based on a number of Chebyshev expansions:
For $x<-5$,
 $Aix=atsin⁡z-btcos⁡z-x1/4$
where $z=\frac{\pi }{4}+\frac{2}{3}\sqrt{-{x}^{3}}$, and $a\left(t\right)$ and $b\left(t\right)$ are expansions in the variable $t=-2{\left(\frac{5}{x}\right)}^{3}-1$.
For $-5\le x\le 0$,
 $Aix=ft-xgt,$
where $f$ and $g$ are expansions in $t=-2{\left(\frac{x}{5}\right)}^{3}-1\text{.}$
For $0,
 $Aix=e-3x/2yt,$
where $y$ is an expansion in $t=4x/9-1$.
For $4.5\le x<9$,
 $Aix=e-5x/2ut,$
where $u$ is an expansion in $t=4x/9-3$.
For $x\ge 9$,
 $Aix=e-zvtx1/4,$
where $z=\frac{2}{3}\sqrt{{x}^{3}}$ and $v$ is an expansion in $t=2\left(\frac{18}{z}\right)-1$.
For , the result is set directly to $\mathrm{Ai}\left(0\right)$. 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 $x<-{\left(\frac{3}{2\epsilon }\right)}^{2/3}$, where $\epsilon$ is the machine precision.
For large positive arguments, where $\mathrm{Ai}$ 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:     $\mathrm{x}$ – double scalar
The argument $x$ of the function.

None.

### Output Parameters

1:     $\mathrm{result}$ – double scalar
The result of the function.
2:     $\mathrm{ifail}$int64int32nag_int scalar
${\mathbf{ifail}}={\mathbf{0}}$ unless the function detects an error (see Error Indicators and Warnings).

## Error Indicators and Warnings

Errors or warnings detected by the function:
${\mathbf{ifail}}=1$
x is too large and positive. On soft failure, the function returns zero.
${\mathbf{ifail}}=2$
x is too large and negative. On soft failure, the function returns zero.
${\mathbf{ifail}}=-99$
${\mathbf{ifail}}=-399$
Your licence key may have expired or may not have been installed correctly.
${\mathbf{ifail}}=-999$
Dynamic memory allocation failed.

## 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, $E$, and the relative error, $\epsilon$, are related in principle to the relative error in the argument, $\delta$, by
 $E≃ x Ai′x δ, ε≃ x Ai′x Aix δ.$
In practice, approximate equality is the best that can be expected. When $\delta$, $\epsilon$ or $E$ is of the order of the machine precision, the errors in the result will be somewhat larger.
For small $x$, errors are strongly damped by the function and hence will be bounded by the machine precision.
For moderate negative $x$, the error behaviour is oscillatory but the amplitude of the error grows like
 $amplitude Eδ ∼x5/4π.$
However the phase error will be growing roughly like $\frac{2}{3}\sqrt{{\left|x\right|}^{3}}$ and hence all accuracy will be lost for large negative arguments due to the impossibility of calculating sin and cos to any accuracy if $\frac{2}{3}\sqrt{{\left|x\right|}^{3}}>\frac{1}{\delta }$.
For large positive arguments, the relative error amplification is considerable:
 $ε δ ∼x3.$
This means a loss of roughly two decimal places accuracy for arguments in the region of $20$. However very large arguments are not possible due to the danger of setting underflow and so the errors are limited in practice.

None.

## Example

This example reads values of the argument $x$ from a file, evaluates the function at each value of $x$ and prints the results.
```function s17ag_example

fprintf('s17ag example results\n\n');

x = [-10   -1    0    1    5   10   20];
n = size(x,2);
result = x;

for j=1:n
[result(j), ifail] = s17ag(x(j));
end

disp('      x         Ai_0(x)');
fprintf('%12.3e%12.3e\n',[x; result]);

s17ag_plot;

function s17ag_plot
x = [-25:0.25:10];
for j = 1:numel(x)
[Ai(j), ifail] = s17ag(x(j));
end

fig1 = figure;
plot(x,Ai,'-r');
xlabel('x');
ylabel('Ai(x)');
title('Airy Function Ai(x)');
axis([-25 10 -0.5 0.6]);

```
```s17ag example results

x         Ai_0(x)
-1.000e+01   4.024e-02
-1.000e+00   5.356e-01
0.000e+00   3.550e-01
1.000e+00   1.353e-01
5.000e+00   1.083e-04
1.000e+01   1.105e-10
2.000e+01   1.692e-27
```