File Coverage

lib/Algorithm/CurveFit/Simple.pm
Criterion Covered Total %
statement 196 208 94.2
branch 61 78 78.2
condition 21 40 52.5
subroutine 16 16 100.0
pod 1 1 100.0
total 295 343 86.0


line stmt bran cond sub pod time code
1             package Algorithm::CurveFit::Simple;
2              
3             # ABSTRACT: Convenience wrapper around Algorithm::CurveFit.
4              
5             our $VERSION = '1.03'; # VERSION 1.03
6              
7 5     5   562395 use strict;
  5         21  
  5         111  
8 5     5   20 use warnings;
  5         8  
  5         120  
9 5     5   1876 use Algorithm::CurveFit;
  5         541876  
  5         205  
10 5     5   38 use Time::HiRes;
  5         8  
  5         42  
11 5     5   406 use JSON::PP;
  5         8  
  5         423  
12              
13             our %STATS_H; # side-products of fit() stored here for profiling purposes
14              
15             BEGIN {
16 5     5   24 require Exporter;
17 5         9 our $VERSION = '1.03';
18 5         90 our @ISA = qw(Exporter);
19 5         8866 our @EXPORT_OK = qw(fit %STATS_H);
20             }
21              
22             # fit() - only public function for this distribution
23             # Given at least parameter "xy", generate a best-fit curve within a time limit.
24             # Output: max deviation, avg deviation, implementation source string (perl or C, for now).
25             # Optional parameters and their defaults:
26             # terms => 3 # number of terms in formula, max is 10
27             # time_limit => 3 # number of seconds to try for better fit
28             # inv => 1 # invert sense of curve-fit, from x->y to y->x
29             # impl_lang => 'perl' # programming language used for output implementation: perl, c
30             # impl_name => 'x2y' # name given to output implementation function
31             sub fit {
32 1     1 1 79 my %p = @_;
33              
34 1         13 my $formula = _init_formula(%p);
35 1         5 my ($xdata, $ydata) = _init_data(%p);
36 1         4 my $parameters = _init_parameters($xdata, $ydata, %p);
37              
38 1         1 my $iter_mode = 'time';
39 1         2 my $time_limit = 3; # sane default?
40 1 50       2 $time_limit = 0.01 if ($time_limit < 0.01);
41 1         2 my $n_iter;
42 1 50       3 if (defined($p{iterations})) {
43 0         0 $iter_mode = 'iter';
44 0   0     0 $n_iter = $p{iterations} || 10000;
45             } else {
46 1   33     4 $time_limit = $p{time_limit} // $time_limit;
47 1         2 $n_iter = 10000 * $time_limit; # will use this to figure out how long it -really- takes.
48             }
49            
50 1         2 my ($n_sec, $params_ar_ar);
51 1 50       3 if ($iter_mode eq 'time') {
52 1         4 ($n_sec, $params_ar_ar) = _try_fit($formula, $parameters, $xdata, $ydata, $n_iter, $p{fitter_class});
53 1         4 $STATS_H{iter_mode} = $iter_mode;
54 1         3 $STATS_H{fit_calib_iter} = $n_iter;
55 1         2 $STATS_H{fit_calib_time} = $n_sec;
56 1         2 $STATS_H{fit_calib_parar} = $params_ar_ar;
57 1         3 $n_iter = int(($time_limit / $n_sec) * $n_iter + 1);
58             }
59              
60 1         5 ($n_sec, $params_ar_ar) = _try_fit($formula, $parameters, $xdata, $ydata, $n_iter, $p{fitter_class});
61 1         4 $STATS_H{fit_iter} = $n_iter;
62 1         3 $STATS_H{fit_time} = $n_sec;
63 1         2 $STATS_H{fit_parar} = $params_ar_ar;
64              
65 1         5 my $coderef = _implement_formula($params_ar_ar, "coderef", "", $xdata, \%p);
66 1         4 my ($max_dev, $avg_dev) = _calculate_deviation($coderef, $xdata, $ydata);
67 1   50     4 my $impl_lang = $p{impl_lang} // 'perl';
68 1         3 $impl_lang = lc($impl_lang);
69 1 50       3 my $impl_name = $p{inv} ? "y2x" : "x2y";
70 1   33     6 $impl_name = $p{impl_name} // $impl_name;
71 1         2 my $impl = $coderef;
72 1 50       5 $impl = _implement_formula($params_ar_ar, $impl_lang, $impl_name, $xdata, \%p) unless($impl_lang eq 'coderef');
73 1         17 return ($max_dev, $avg_dev, $impl);
74             }
75              
76             # ($n_sec, $params_ar_ar) = _try_fit($formula, $parameters, $xdata, $ydata, $n_iter, $p{fitter_class});
77             sub _try_fit {
78 2     2   9 my ($formula, $parameters, $xdata, $ydata, $n_iter, $fitter_class) = @_;
79 2   50     18 $fitter_class //= "Algorithm::CurveFit";
80 2         4 my $params_ar_ar = [map {[@$_]} @$parameters]; # making a copy because curve_fit() is destructive
  8         15  
81 2         6 my $tm0 = Time::HiRes::time();
82 2         16 my $res = $fitter_class->curve_fit(
83             formula => $formula,
84             params => $params_ar_ar,
85             variable => 'x',
86             xdata => $xdata,
87             ydata => $ydata,
88             maximum_iterations => $n_iter
89             );
90 2         203406 my $tm_elapsed = Time::HiRes::time() - $tm0;
91 2         9 return ($tm_elapsed, $params_ar_ar);
92             }
93              
94             sub _init_formula {
95 11     11   10092 my %p = @_;
96 11         17 my $formula = 'k + a*x + b*x^2 + c*x^3'; # sane'ish default
97 11   100     41 my $terms = $p{terms} // 3;
98 11 50       20 die "maximum of 10 terms\n" if ($terms > 10);
99 11 100       20 if ($terms != 3) {
100 8         10 $formula = 'k';
101 8         14 for (my $i = 1; $i <= $terms; $i++) {
102 42         44 my $fact = chr(ord('a') + $i - 1);
103 42         79 $formula .= " + $fact * x^$i";
104             }
105             }
106 11         48 return $formula;
107             }
108              
109             # ($xdata, $ydata) = _init_data(%p);
110             sub _init_data {
111 15     15   10628 my %p = @_;
112 15         24 my ($xdata, $ydata);
113 15 100 100     66 if (defined($p{xydata})) {
    100          
114 7         9 my $xy = $p{xydata};
115 7 50 33     54 unless (
      33        
      33        
116             ref($xy) eq 'ARRAY'
117             && @$xy >= 2
118             && ref($xy->[0]) eq 'ARRAY'
119             && ref($xy->[1]) eq 'ARRAY'
120             ) {
121 0         0 die "xydata must be either an arrayref of [x, y] data point arrayrefs or an arrayref [[x0, x1, ... xN], [y0, y1, ... yN]]\n";
122             }
123 7 100 100     17 if (@$xy == 2 && @{$xy->[0]} > 2) {
  3         7  
124             # user has provided [[x, ..], [y, ..]]
125 2         2 $xdata = $xy->[0];
126 2         3 $ydata = $xy->[1];
127             } else {
128             # user has provided [[x, y], [x, y], ..]
129 5 100       6 die "pairwise xydata must have two data points per element\n" unless(@{$xy->[0]} == 2);
  5         16  
130 4         6 $xdata = [map {$_->[0]} @{$xy}];
  21         26  
  4         7  
131 4         6 $ydata = [map {$_->[1]} @{$xy}];
  21         21  
  4         5  
132             }
133             }
134             elsif (defined($p{xdata}) && defined($p{ydata})) {
135 5         8 $xdata = $p{xdata};
136 5         4 $ydata = $p{ydata};
137             }
138             else {
139 3         14 die "must provide at least xydata or both xdata and ydata\n";
140             }
141 11 50 33     40 die "xdata and ydata must both be arrayref\n" unless (ref($xdata) eq "ARRAY" && ref($ydata) eq "ARRAY");
142 11 100       31 die "xdata and ydata must have the same number of elements\n" unless (@$xdata == @$ydata);
143 9 100       17 die "must have more than one data-point\n" unless (@$xdata > 1);
144 8 100       15 if ($p{inv}) {
145 3         5 $STATS_H{xdata} = $ydata;
146 3         4 $STATS_H{ydata} = $xdata;
147 3         18 return ($ydata, $xdata);
148             }
149 5         8 $STATS_H{xdata} = $xdata;
150 5         7 $STATS_H{ydata} = $ydata;
151 5         16 return ($xdata, $ydata);
152             }
153              
154             # $parameters = _init_parameters($xdata, $ydata, %p);
155             sub _init_parameters {
156 3     3   1207 my ($xdata, $ydata, %p) = @_;
157 3         5 my $k = 0;
158 3         4 my $n_points = @$xdata;
159 3         6 foreach my $v (@$ydata) { $k += $v; }
  18         18  
160 3         5 $k /= $n_points;
161             # zzapp -- implement any precision hints here.
162 3         9 my @params = (['k', $k, 0.0000001]);
163 3   100     11 my $terms = $p{terms} // 3;
164 3         8 push @params, map {[chr(ord('a')+$_-1), 0.5, 0.0000001]} (1..$terms);
  8         18  
165 3         10 return \@params;
166             }
167              
168             # $impl = _implement_formula($params_ar_ar, $impl_lang, $impl_name, $xdata, \%p) unless($impl_lang eq 'coderef');
169             sub _implement_formula {
170 10     10   4609 my ($params_ar_ar, $impl_lang, $impl_name, $xdata, $opt_hr) = @_;
171 10 100       28 return _implement_formula_as_coderef(@_) if ($impl_lang eq 'coderef');
172             # return _implement_formula_as_python(@_) if ($impl_lang eq 'python'); # zzapp
173 8 100       18 return _implement_formula_as_C(@_) if ($impl_lang eq 'c');
174             # return _implement_formula_as_R(@_) if ($impl_lang eq 'r'); # zzapp
175             # return _implement_formula_as_MATLAB(@_) if ($impl_lang eq 'matlab'); # zzapp
176 5         14 return _implement_formula_as_perl(@_);
177             }
178              
179             sub _implement_formula_as_coderef {
180 2     2   6 my ($params_ar_ar, $impl_lang, $impl_name, $xdata, $opt_hr) = @_;
181 2         4 my $k_ar = $params_ar_ar->[0];
182 2         15 my $formula = sprintf("%f", $k_ar->[1]);
183 2         6 for (my $i = 1; defined($params_ar_ar->[$i]); $i++) {
184 6         9 my $fact = $params_ar_ar->[$i]->[1];
185 6 100       14 my $pow = ($i == 1) ? "" : "**$i";
186 6         26 $formula .= sprintf(' + %f * $x%s', $fact, $pow);
187             }
188 2         4 $STATS_H{impl_formula} = $formula;
189 2         5 my $bounder = '';
190 2 50       8 if ($opt_hr->{bounds_check}) {
191 0         0 my ($high_x, $low_x) = ($xdata->[0], $xdata->[0]);
192 0         0 foreach my $x (@$xdata) {
193 0 0       0 $high_x = $x if ($high_x < $x);
194 0 0       0 $low_x = $x if ($low_x > $x);
195             }
196 0         0 $bounder = 'die "x out of bounds (high)" if ($x > '.$high_x.'); die "x out of bounds (low)" if ($x < '.$low_x.');';
197             }
198 2         3 my $rounder = '';
199 2 50       6 $rounder = '$y = int($y + 0.5);' if ($opt_hr->{round_result});
200 2         8 my $src = 'sub { my($x) = @_; '.$bounder.' my $y = '.$formula.'; '.$rounder.' return $y; }';
201 2         4 $STATS_H{impl_source} = $src;
202 2         3 $STATS_H{impl_exception} = '';
203 2         195 my $coderef = eval($src);
204 2 50       8 $STATS_H{impl_exception} = $@ unless(defined($coderef));
205 2         7 return $coderef;
206             }
207              
208             sub _implement_formula_as_perl {
209 5     5   8 my ($params_ar_ar, $impl_lang, $impl_name, $xdata, $opt_hr) = @_;
210 5         9 my $k_ar = $params_ar_ar->[0];
211 5         31 my $formula = sprintf("%.11f", $k_ar->[1]);
212 5         14 for (my $i = 1; defined($params_ar_ar->[$i]); $i++) {
213 15         17 my $fact = $params_ar_ar->[$i]->[1];
214 15 100       31 my $pow = ($i == 1) ? "" : "**$i";
215 15         54 $formula .= sprintf(' + %.11f * $x%s', $fact, $pow);
216             }
217 5         18 $STATS_H{impl_formula} = $formula;
218 5         6 my $bounder = '';
219 5 100       12 if ($opt_hr->{bounds_check}) {
220 1         3 my ($high_x, $low_x) = ($xdata->[0], $xdata->[0]);
221 1         2 foreach my $x (@$xdata) {
222 7 100       9 $high_x = $x if ($high_x < $x);
223 7 100       10 $low_x = $x if ($low_x > $x);
224             }
225 1         7 $bounder = sprintf(' die "x out of bounds (high)" if ($x > %.11f);'."\n", $high_x) .
226             sprintf(' die "x out of bounds (low)" if ($x < %.11f);'."\n", $low_x);
227             }
228 5         8 my $rounder = '';
229 5 100       14 $rounder = ' $y = int($y + 0.5);'."\n" if ($opt_hr->{round_result});
230 5         21 my $src = join("\n",(
231             "sub $impl_name {",
232             ' my($x) = @_;',
233             $bounder,
234             ' my $y = '.$formula.';',
235             $rounder,
236             ' return $y;',
237             '}'
238             ));
239 5         8 $STATS_H{impl_source} = $src;
240 5         7 $STATS_H{impl_exception} = '';
241 5         14 return $src;
242             }
243              
244             sub _implement_formula_as_C {
245 3     3   18 my ($params_ar_ar, $impl_lang, $impl_name, $xdata, $opt_hr) = @_;
246 3         5 my $k_ar = $params_ar_ar->[0];
247 3         3 my $src = "";
248 3 100 66     10 $src .= "#include \n" if ($opt_hr->{round_result} && !$opt_hr->{suppress_includes});
249 3         6 $src .= "double $impl_name(double x) {\n";
250 3         16 $src .= sprintf(" double y = %.11f;\n", $k_ar->[1]);
251 3         5 $src .= " double xx = x;\n"; # eliminating pow() calls, which gcc doesn't seem willing to optimize completely away
252              
253 3 100       7 if ($opt_hr->{bounds_check}) {
254 1         2 my ($high_x, $low_x) = ($xdata->[0], $xdata->[0]);
255 1         3 foreach my $x (@$xdata) {
256 7 100       11 $high_x = $x if ($high_x < $x);
257 7 100       8 $low_x = $x if ($low_x > $x);
258             }
259             # zzapp -- this is kludgy. better way to signal bounds violation?
260 1         8 $src .= sprintf(" if (x > %.11f) return -1.0;\n", $high_x) .
261             sprintf(" if (x < %.11f) return -1.0;\n", $low_x);
262             }
263              
264 3         4 my $formula = "";
265 3         8 for (my $i = 1; defined($params_ar_ar->[$i]); $i++) {
266 9         10 my $fact = $params_ar_ar->[$i]->[1];
267 9         21 $formula .= sprintf(" y += %.11f * xx;\n", $fact);
268 9 100       23 $formula .= " xx *= x;\n" if(defined($params_ar_ar->[$i+1]));
269             }
270 3         4 $STATS_H{impl_formula} = $formula; # zzapp -- not clean!
271              
272 3         5 $src .= $formula;
273 3 100       7 $src .= " y = round(y);\n" if ($opt_hr->{round_result});
274 3         4 $src .= " return y;\n}\n";
275 3         3 $STATS_H{impl_source} = $src;
276 3         4 $STATS_H{impl_exception} = '';
277 3         8 return $src;
278             }
279              
280             # ($max_dev, $avg_dev) = _calculate_deviation($coderef, $xdata, $ydata);
281             sub _calculate_deviation {
282 1     1   2 my ($coderef, $xdata, $ydata) = @_;
283 1         2 my $max_off = 1.0;
284 1         2 my $max_off_datum = 0.0;
285 1         2 my $tot_off = 0.0;
286 1         8 for (my $i = 0; defined($xdata->[$i]); $i++) {
287 10         9 my $x = $xdata->[$i];
288 10         8 my $y = eval { $coderef->($x); };
  10         142  
289 10 50       39 unless(defined($y)) {
290 0         0 $STATS_H{deviation_exception} = $@;
291 0         0 $STATS_H{deviation_exception_datum} = $x;
292 0         0 die "caught exception calculating deviations";
293             }
294              
295 10         14 my $observed_y = $ydata->[$i];
296 10 50 33     24 if ($observed_y && $y) {
297 10 100       17 my $deviation = $y > $observed_y ? $y / $observed_y : $observed_y / $y;
298 10         11 my $dev_mag = abs($deviation - 1.0);
299 10         11 my $max_mag = abs($max_off - 1.0);
300             # print "x=$x\ty=$y\toy=$observed_y\tdev_mag=$dev_mag\tmax_mag=$max_mag\n";
301 10 100       15 ($max_off, $max_off_datum) = ($deviation, $x) if ($dev_mag > $max_mag);
302 10         19 $tot_off += $deviation;
303             } else {
304 0         0 $tot_off += 1.0;
305             }
306             }
307 1         3 $STATS_H{deviation_max_offset_datum} = $max_off_datum;
308 1         4 return ($max_off, $tot_off / @$xdata);
309             }
310              
311              
312             1;
313              
314             =head1 NAME
315              
316             Algorithm::CurveFit::Simple - Convenience wrapper around Algorithm::CurveFit
317              
318             =head1 SYNOPSIS
319              
320             use Algorithm::CurveFit::Simple qw(fit);
321              
322             my ($max_dev, $avg_dev, $src) = fit(xdata => \@xdata, ydata => \@ydata, ..options..);
323              
324             # Alternatively pass xdata and ydata together:
325             my ($max_dev, $avg_dev, $src) = fit(xydata => [\@xdata, \@ydata], ..options..);
326              
327             # Alternatively pass data as array of [x,y] pairs:
328             my ($max_dev, $avg_dev, $src) = fit(xydata => [[1, 2], [2, 5], [3, 10]], ..options..);
329              
330             =head1 DESCRIPTION
331              
332             This is a convenience wrapper around L. Given a body of (x, y) data points, it will generate a polynomial formula f(x) = y which fits that data.
333              
334             Its main differences from L are:
335              
336             =over 4
337              
338             =item * It synthesizes the initial formula for you,
339              
340             =item * It allows for a time limit on the curve-fit instead of an iteration count,
341              
342             =item * It implements the formula as source code (or as a perl coderef, if you want to use the formula immediately in your program).
343              
344             =back
345              
346             Additionally it returns a maximum deviation and average deviation of the formula vs the xydata, which is more useful (to me, at least) than L's square residual output. Closer to 1.0 indicates a better fit. Play with C #> until these deviations are as close to 1.0 as possible, and beware overfitting.
347              
348             =head1 SUBROUTINES
349              
350             There is only one public subroutine, C. It B be given either C or C and C parameters. All other paramters are optional.
351              
352             It returns three values: A maximum deviation, the average deviation and the formula implementation.
353              
354             =head2 Options
355              
356             =over 4
357              
358             =item C \@xdata, ydata =E \@ydata)>
359              
360             The data points the formula will fit. Same as L parameters of the same name.
361              
362             =item C [[1, 2, 3, 4], [10, 17, 26, 37]])>
363              
364             =item C [[1, 10], [2, 17], [3, 26], [4, 37]])>
365              
366             A more convenient way to provide data points. C will try to detect how the data points are organized -- list of x and list of y, or list of [x,y].
367              
368             =item C 3)>
369              
370             Sets the order of the polynomial, which will be of the form C. The default is 3 and the limit is 10.
371              
372             There is no need to specify initial C. It will be calculated from C.
373              
374             =item C 3)>
375              
376             If a time limit is given (in seconds), C will spend no more than that long trying to fit the data. It may return in much less time. The default is 3.
377              
378             =item C 10000)>
379              
380             If an iteration count is given, C will ignore any time limit and iterate up to C times trying to fit the curve. Same as L parameter of the same name.
381              
382             =item C 1)>
383              
384             Setting C inverts the sense of the fit. Instead of C the formula will fit C.
385              
386             =item C "perl")>
387              
388             Sets the programming language in which the formula will be implemented. Currently supported languages are C<"C">, C<"coderef"> and the default, C<"perl">.
389              
390             When C "coderef"> is specified, a code reference is returned instead which may be used immediately by your perl script:
391              
392             my($max_dev, $avg_dev, $x2y) = fit(xydata => \@xy, impl_lang => "coderef");
393              
394             my $y = $x2y->(42);
395              
396             More implementation languages will be supported in the future.
397              
398             =item C "x2y")>
399              
400             Sets the name of the function implementing the formula. The default is C<"x2y">. Has no effect when used with C "coderef")>.
401              
402             my($max_dev, $avg_dev, $src) = fit(xydata => \@xy, impl_name => "converto");
403              
404             print "$src\n";
405              
406             sub converto {
407             my($x) = @_;
408             my $y = -5340.93059104837 + 249.23009968947 * $x + -3.87745746448 * $x**2 + 0.02114780993 * $x**3;
409             return $y;
410             }
411              
412             =item C 1)>
413              
414             When set, the implementation will include logic for checking whether the input is out-of-bounds, per the highest and lowest x points in the data used to fit the formula. For implementation languages which support exceptions, an exception will be thrown. For others (like C), C<-1.0> will be returned to indicate the error.
415              
416             For instance, if the highest x in C<$xydata> is 83.0 and the lowest x is 60.0:
417              
418             my($max_dev, $avg_dev, $src) = fit(xydata => \@xy, bounds_check => 1);
419              
420             print "$src\n";
421              
422             sub x2y {
423             my($x) = @_;
424             die "x out of bounds (high)" if ($x > 83.80000000000);
425             die "x out of bounds (low)" if ($x < 60.80000000000);
426             my $y = -5340.93059104837 + 249.23009968947 * $x + -3.87745746448 * $x**2 + 0.02114780993 * $x**3;
427             return $y;
428             }
429              
430             =item C 1)>
431              
432             When set, the implementation will round the output to the nearest whole number. When the implementation language is C<"C"> this adds an C<#include Emath.hE> directive to the source code, which will have to be compiled against libm -- see C.
433              
434             my($max_dev, $avg_dev, $src) = fit(xydata => \@xy, round_result => 1);
435              
436             print "$src\n";
437              
438             sub x2y {
439             my($x) = @_;
440             my $y = -5340.93059104837 + 249.23009968947 * $x + -3.87745746448 * $x**2 + 0.02114780993 * $x**3;
441             $y = int($y + 0.5);
442             return $y;
443             }
444              
445             =item C 1)>
446              
447             When set and C "C">, any C<#include> directives which the implementation might need will be suppressed.
448              
449             =back
450              
451             =head1 VARIABLES
452              
453             The class variable C<%STATS_H> contains various intermediate values which might be helpful. For instance, C<$STATS_H{deviation_max_offset_datum}> contains the x data point which corresponds to the maximum deviation returned.
454              
455             The contents of C<%STATS_H> is subject to change and might not be fully documented in future versions. The current fields are:
456              
457             =over 4
458              
459             =item C: The x data point corresponding with returned maximum deviation.
460              
461             =item C: Arrayref of formula parameters as returned by L after a short fitting attempt used for timing calibration.
462              
463             =item C: The number of seconds L spent in the calibration run.
464              
465             =item C: The iterations parameter passed to L.
466              
467             =item C: Arrayref of formula parameters as returned by L.
468              
469             =item C: The number of seconds L actually spent fitting the formula.
470              
471             =item C: The exception thrown when the implementation was used to calculate the deviations, or the empty string if none.
472              
473             =item C: The formula part of the implementation.
474              
475             =item C: The implementation source string.
476              
477             =item C: One of C<"time"> or C<"iter">, indicating whether a time limit was used or an iteration count.
478              
479             =item C: Arrayref of x data points as passed to L.
480              
481             =item C: Arrayref of y data points as passed to L.
482              
483             =back
484              
485             =head1 CAVEATS
486              
487             =over 4
488              
489             =item * Only simple polynomial functions are supported. Sometimes you need something else. Use L for such cases.
490              
491             =item * If C is very large, iterating over it to calculate deviances can take more time than permitted by C.
492              
493             =item * The dangers of overfitting are real! L
494              
495             =item * Using too many terms can dramatically reduce the accuracy of the fitted formula.
496              
497             =item * Sometimes calling L with a ten-term polynomial causes it to hang.
498              
499             =back
500              
501             =head1 TO DO
502              
503             =over 4
504              
505             =item * Support more programming languages for formula implementation: R, MATLAB, python
506              
507             =item * Calculate the actual term sigfigs and set precision appropriately in the formula implementation instead of just "%.11f".
508              
509             =item * Support trying a range of terms and returning whatever gives the best fit.
510              
511             =item * Support piecewise output formulas.
512              
513             =item * Work around L's occasional hang problem when using ten-term polynomials.
514              
515             =back
516              
517             =head1 SEE ALSO
518              
519             L
520              
521             L
522              
523             =cut