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_smooth_fit_spline_parest (g10ac)

## Purpose

nag_smooth_fit_spline_parest (g10ac) estimates the values of the smoothing parameter and fits a cubic smoothing spline to a set of data.

## Syntax

[yhat, c, rss, df, res, h, crit, rho, ifail] = g10ac(method, x, y, crit, 'n', n, 'wt', wt, 'u', u, 'tol', tol, 'maxcal', maxcal)
[yhat, c, rss, df, res, h, crit, rho, ifail] = nag_smooth_fit_spline_parest(method, x, y, crit, 'n', n, 'wt', wt, 'u', u, 'tol', tol, 'maxcal', maxcal)
Note: the interface to this routine has changed since earlier releases of the toolbox:
 At Mark 24: tol was made optional; weight was removed from the interface; wt was made optional

## Description

For a set of $n$ observations $\left({x}_{\mathit{i}},{y}_{\mathit{i}}\right)$, for $\mathit{i}=1,2,\dots ,n$, the spline provides a flexible smooth function for situations in which a simple polynomial or nonlinear regression model is not suitable.
Cubic smoothing splines arise as the unique real-valued solution function $f$, with absolutely continuous first derivative and squared-integrable second derivative, which minimizes
 $∑i=1nwiyi-fxi2+ρ ∫-∞∞f′′x2dx,$
where ${w}_{i}$ is the (optional) weight for the $i$th observation and $\rho$ is the smoothing argument. This criterion consists of two parts: the first measures the fit of the curve and the second the smoothness of the curve. The value of the smoothing argument $\rho$ weights these two aspects; larger values of $\rho$ give a smoother fitted curve but, in general, a poorer fit. For details of how the cubic spline can be fitted see Hutchinson and de Hoog (1985) and Reinsch (1967).
The fitted values, $\stackrel{^}{y}={\left({\stackrel{^}{y}}_{1},{\stackrel{^}{y}}_{2},\dots ,{\stackrel{^}{y}}_{n}\right)}^{\mathrm{T}}$, and weighted residuals, ${r}_{i}$, can be written as:
 $y^=Hy and ri=wiyi-y^i$
for a matrix $H$. The residual degrees of freedom for the spline is $\mathrm{trace}\left(I-H\right)$ and the diagonal elements of $H$ are the leverages.
The parameter $\rho$ can be estimated in a number of ways.
(i) The degrees of freedom for the spline can be specified, i.e., find $\rho$ such that $\mathrm{trace}\left(H\right)={\nu }_{0}$ for given ${\nu }_{0}$.
(ii) Minimize the cross-validation (CV), i.e., find $\rho$ such that the CV is minimized, where
 $CV=1∑i=1nwi ∑i=1n ri1-hii 2.$
(iii) Minimize the generalized cross-validation (GCV), i.e., find $\rho$ such that the GCV is minimized, where
 $GCV=n2∑i=1nwi ∑i=1nri2 ∑i=1n1-hii 2 .$
nag_smooth_fit_spline_parest (g10ac) requires the ${x}_{i}$ to be strictly increasing. If two or more observations have the same ${x}_{i}$ value then they should be replaced by a single observation with ${y}_{i}$ equal to the (weighted) mean of the $y$ values and weight, ${w}_{i}$, equal to the sum of the weights. This operation can be performed by nag_smooth_data_order (g10za)
The algorithm is based on Hutchinson (1986). nag_roots_contfn_brent_rcomm (c05az) is used to solve for $\rho$ given ${\nu }_{0}$ and the method of nag_opt_one_var_func (e04ab) is used to minimize the GCV or CV.

## References

Hastie T J and Tibshirani R J (1990) Generalized Additive Models Chapman and Hall
Hutchinson M F (1986) Algorithm 642: A fast procedure for calculating minimum cross-validation cubic smoothing splines ACM Trans. Math. Software 12 150–153
Hutchinson M F and de Hoog F R (1985) Smoothing noisy data with spline functions Numer. Math. 47 99–106
Reinsch C H (1967) Smoothing by spline functions Numer. Math. 10 177–183

## Parameters

### Compulsory Input Parameters

1:     $\mathrm{method}$ – string (length ≥ 1)
Indicates whether the smoothing parameter is to be found by minimization of the CV or GCV functions, or by finding the smoothing parameter corresponding to a specified degrees of freedom value.
${\mathbf{method}}=\text{'C'}$
Cross-validation is used.
${\mathbf{method}}=\text{'D'}$
The degrees of freedom are specified.
${\mathbf{method}}=\text{'G'}$
Generalized cross-validation is used.
Constraint: ${\mathbf{method}}=\text{'C'}$, $\text{'D'}$ or $\text{'G'}$.
2:     $\mathrm{x}\left({\mathbf{n}}\right)$ – double array
The distinct and ordered values ${x}_{\mathit{i}}$, for $\mathit{i}=1,2,\dots ,n$.
Constraint: ${\mathbf{x}}\left(\mathit{i}\right)<{\mathbf{x}}\left(\mathit{i}+1\right)$, for $\mathit{i}=1,2,\dots ,n-1$.
3:     $\mathrm{y}\left({\mathbf{n}}\right)$ – double array
The values ${y}_{\mathit{i}}$, for $\mathit{i}=1,2,\dots ,n$.
4:     $\mathrm{crit}$ – double scalar
If ${\mathbf{method}}=\text{'D'}$, the required degrees of freedom for the spline.
If ${\mathbf{method}}=\text{'C'}$ or $\text{'G'}$, crit need not be set.
Constraint: $2.0<{\mathbf{crit}}\le {\mathbf{n}}$.

### Optional Input Parameters

1:     $\mathrm{n}$int64int32nag_int scalar
Default: the dimension of the arrays x, y. (An error is raised if these dimensions are not equal.)
$n$, the number of observations.
Constraint: ${\mathbf{n}}\ge 3$.
2:     $\mathrm{wt}\left(:\right)$ – double array
The dimension of the array wt must be at least ${\mathbf{n}}$ if $\mathit{weight}=\text{'W'}$
If $\mathit{weight}=\text{'W'}$, wt must contain the $n$ weights. Otherwise wt is not referenced and unit weights are assumed.
Constraint: if $\mathit{weight}=\text{'W'}$, ${\mathbf{wt}}\left(\mathit{i}\right)>0.0$, for $\mathit{i}=1,2,\dots ,n$.
3:     $\mathrm{u}$ – double scalar
Default: $0.0$
The upper bound on the smoothing parameter. If ${\mathbf{u}}\le {\mathbf{tol}}$, ${\mathbf{u}}=1000.0$ will be used instead. See Further Comments for details on how this argument is used.
4:     $\mathrm{tol}$ – double scalar
Default: $0.0$
The accuracy to which the smoothing parameter rho is required. tol should preferably be not much less than $\sqrt{\epsilon }$, where $\epsilon$ is the machine precision. If ${\mathbf{tol}}<\epsilon$, ${\mathbf{tol}}=\sqrt{\epsilon }$ will be used instead.
5:     $\mathrm{maxcal}$int64int32nag_int scalar
Default: $0$
The maximum number of spline evaluations to be used in finding the value of $\rho$. If ${\mathbf{maxcal}}<3$, ${\mathbf{maxcal}}=100$ will be used instead.

### Output Parameters

1:     $\mathrm{yhat}\left({\mathbf{n}}\right)$ – double array
The fitted values, ${\stackrel{^}{y}}_{\mathit{i}}$, for $\mathit{i}=1,2,\dots ,n$.
2:     $\mathrm{c}\left(\mathit{ldc},3\right)$ – double array
The spline coefficients. More precisely, the value of the spline approximation at $t$ is given by $\left(\left({\mathbf{c}}\left(i,3\right)×d+{\mathbf{c}}\left(i,2\right)\right)×d+{\mathbf{c}}\left(i,1\right)\right)×d+{\stackrel{^}{y}}_{i}$, where ${x}_{i}\le t<{x}_{i+1}$ and $d=t-{x}_{i}$.
3:     $\mathrm{rss}$ – double scalar
The (weighted) residual sum of squares.
4:     $\mathrm{df}$ – double scalar
The residual degrees of freedom. If ${\mathbf{method}}=\text{'D'}$ this will be $n-{\mathbf{crit}}$ to the required accuracy.
5:     $\mathrm{res}\left({\mathbf{n}}\right)$ – double array
The (weighted) residuals, ${r}_{\mathit{i}}$, for $\mathit{i}=1,2,\dots ,n$.
6:     $\mathrm{h}\left({\mathbf{n}}\right)$ – double array
The leverages, ${h}_{\mathit{i}\mathit{i}}$, for $\mathit{i}=1,2,\dots ,n$.
7:     $\mathrm{crit}$ – double scalar
If ${\mathbf{method}}=\text{'C'}$, the value of the cross-validation, or if ${\mathbf{method}}=\text{'G'}$, the value of the generalized cross-validation function, evaluated at the value of $\rho$ returned in rho.
8:     $\mathrm{rho}$ – double scalar
The smoothing parameter, $\rho$.
9:     $\mathrm{ifail}$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:

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

${\mathbf{ifail}}=1$
Constraint: if ${\mathbf{method}}=\text{'D'}$, ${\mathbf{crit}}>2.0$.
Constraint: if ${\mathbf{method}}=\text{'D'}$, ${\mathbf{crit}}\le {\mathbf{n}}$.
Constraint: $\mathit{ldc}\ge {\mathbf{n}}-1$.
Constraint: ${\mathbf{n}}\ge 3$.
On entry, method is not valid.
${\mathbf{ifail}}=2$
On entry, at least one element of ${\mathbf{wt}}\le 0.0$.
${\mathbf{ifail}}=3$
On entry, x is not a strictly ordered array.
${\mathbf{ifail}}=4$
For the specified degrees of freedom, ${\mathbf{rho}}>{\mathbf{u}}$:
W  ${\mathbf{ifail}}=5$
Accuracy of tol cannot be achieved:
W  ${\mathbf{ifail}}=6$
maxcal iterations have been performed.
W  ${\mathbf{ifail}}=7$
Optimum value of rho lies above u:
${\mathbf{ifail}}=-99$
${\mathbf{ifail}}=-399$
Your licence key may have expired or may not have been installed correctly.
${\mathbf{ifail}}=-999$
Dynamic memory allocation failed.

## Accuracy

When minimizing the cross-validation or generalized cross-validation, the error in the estimate of $\rho$ should be within $±3\left({\mathbf{tol}}×{\mathbf{rho}}+{\mathbf{tol}}\right)$. When finding $\rho$ for a fixed number of degrees of freedom the error in the estimate of $\rho$ should be within $±2×{\mathbf{tol}}×\mathrm{max}\phantom{\rule{0.125em}{0ex}}\left(1,{\mathbf{rho}}\right)$.
Given the value of $\rho$, the accuracy of the fitted spline depends on the value of $\rho$ and the position of the $x$ values. The values of ${x}_{i}-{x}_{i-1}$ and ${w}_{i}$ are scaled and $\rho$ is transformed to avoid underflow and overflow problems.

The time to fit the spline for a given value of $\rho$ is of order $n$.
When finding the value of $\rho$ that gives the required degrees of freedom, the algorithm examines the interval $0.0$ to u. For small degrees of freedom the value of $\rho$ can be large, as in the theoretical case of two degrees of freedom when the spline reduces to a straight line and $\rho$ is infinite. If the CV or GCV is to be minimized then the algorithm searches for the minimum value in the interval $0.0$ to u. If the function is decreasing in that range then the boundary value of u will be returned. In either case, the larger the value of u the more likely is the interval to contain the required solution, but the process will be less efficient.
Regression splines with a small $\left( number of knots can be fitted by nag_fit_1dspline_knots (e02ba) and nag_fit_1dspline_auto (e02be).

## Example

This example uses the data given by Hastie and Tibshirani (1990), which consists of the age, ${x}_{i}$, and C-peptide concentration (pmol/ml), ${y}_{i}$, from a study of the factors affecting insulin-dependent diabetes mellitus in children. The data is input, reduced to a strictly ordered set by nag_smooth_data_order (g10za) and a spline with $5$ degrees of freedom is fitted by nag_smooth_fit_spline_parest (g10ac). The fitted values and residuals are printed.
```function g10ac_example

fprintf('g10ac example results\n\n');

x =  [ 5.2  8.8 10.5 10.6 10.4  1.8 12.7 15.6  5.8  1.9 ...
2.2  4.8  7.9  5.2  0.9 11.8  7.9 11.5 10.6  8.5 ...
11.1 12.8 11.3  1.0 14.5 11.9  8.1 13.8 15.5  9.8 ...
11.0 12.4 11.1  5.1  4.8  4.2  6.9 13.2  9.9 12.5 ...
13.2  8.9 10.8];
y =  [ 4.8  4.1  5.2  5.5  5.0  3.4  3.4  4.9  5.6  3.7 ...
3.9  4.5  4.8  4.9  3.0  4.6  4.8  5.5  4.5  5.3 ...
4.7  6.6  5.1  3.9  5.7  5.1  5.2  3.7  4.9  4.8 ...
4.4  5.2  5.1  4.6  3.9  5.1  5.1  6.0  4.9  4.1 ...
4.6  4.9  5.1];

% Reorder x, remove ties and weight accordingly
[n, x, y, wt, rss, ifail] = g10za( ...
x, y);
x = x(1:n);
y = y(1:n);

% Control parameters
crit = 12;

% fit cubic spline
method = 'D';
[yhat, c, rss, df, res, h, crit, rho, ifail] = ...
g10ac( ...
method, x, y, crit, 'wt', wt);

%  Display results
fprintf('Residual sum of squares     = %10.2f\n', rss);
fprintf('Degrees of freedom          = %10.2f\n', df);
fprintf('rho                         = %10.2f\n', rho);
fprintf('\n     Input data                Output results\n');
fprintf('   i     x       y            yhat      h\n');
ivar = double(1:n)';
fprintf('%4d%8.3f%8.3f%14.3f%8.3f\n', [ivar x y yhat h]');

```
```g10ac example results

Residual sum of squares     =      10.35
Degrees of freedom          =      25.00
rho                         =       2.68

Input data                Output results
i     x       y            yhat      h
1   0.900   3.000         3.373   0.534
2   1.000   3.900         3.406   0.427
3   1.800   3.400         3.642   0.313
4   1.900   3.700         3.686   0.313
5   2.200   3.900         3.839   0.448
6   4.200   5.100         4.614   0.564
7   4.800   4.200         4.576   0.442
8   5.100   4.600         4.715   0.189
9   5.200   4.850         4.783   0.407
10   5.800   5.600         5.193   0.455
11   6.900   5.100         5.184   0.592
12   7.900   4.800         4.958   0.530
13   8.100   5.200         4.931   0.235
14   8.500   5.300         4.845   0.245
15   8.800   4.100         4.763   0.271
16   8.900   4.900         4.748   0.292
17   9.800   4.800         4.850   0.301
18   9.900   4.900         4.875   0.277
19  10.400   5.000         4.970   0.173
20  10.500   5.200         4.977   0.154
21  10.600   5.000         4.979   0.285
22  10.800   5.100         4.970   0.136
23  11.000   4.400         4.961   0.137
24  11.100   4.900         4.964   0.284
25  11.300   5.100         4.975   0.162
26  11.500   5.500         4.975   0.186
27  11.800   4.600         4.930   0.213
28  11.900   5.100         4.911   0.220
29  12.400   5.200         4.852   0.206
30  12.500   4.100         4.857   0.196
31  12.700   3.400         4.900   0.189
32  12.800   6.600         4.932   0.193
33  13.200   5.300         4.955   0.488
34  13.800   3.700         4.797   0.408
35  14.500   5.700         5.076   0.559
36  15.500   4.900         4.979   0.445
37  15.600   4.900         4.946   0.535
```