| 1 | # Copyright 2007, 2009, 2010, 2011 Kevin Ryde | ||
| 2 | |||
| 3 | # This file is part of Chart. | ||
| 4 | # | ||
| 5 | # Chart is free software; you can redistribute it and/or modify it under the | ||
| 6 | # terms of the GNU General Public License as published by the Free Software | ||
| 7 | # Foundation; either version 3, or (at your option) any later version. | ||
| 8 | # | ||
| 9 | # Chart is distributed in the hope that it will be useful, but WITHOUT ANY | ||
| 10 | # WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS | ||
| 11 | # FOR A PARTICULAR PURPOSE. See the GNU General Public License for more | ||
| 12 | # details. | ||
| 13 | # | ||
| 14 | # You should have received a copy of the GNU General Public License along | ||
| 15 | # with Chart. If not, see <http://www.gnu.org/licenses/>. | ||
| 16 | |||
| 17 | package App::Chart::Series::Derived::AdaptiveLaguerre; | ||
| 18 | use 5.010; | ||
| 19 | use strict; | ||
| 20 | use warnings; | ||
| 21 | use Carp; | ||
| 22 | use List::Util qw(min max); | ||
| 23 | use Locale::TextDomain ('App-Chart'); | ||
| 24 | |||
| 25 | use base 'App::Chart::Series::Indicator'; | ||
| 26 | use App::Chart::Series::Derived::LaguerreFilter; | ||
| 27 | use App::Chart::Series::Derived::Median; | ||
| 28 | use App::Chart::Series::Derived::WilliamsR; | ||
| 29 | |||
| 30 | |||
| 31 | # http://www.mesasoftware.com/technicalpapers.htm | ||
| 32 | # http://www.mesasoftware.com/Papers/TIME%20WARP.pdf | ||
| 33 | # Paper by John Elhers. | ||
| 34 | # | ||
| 35 | # http://www.mesasoftware.com/seminars.htm | ||
| 36 | # http://www.mesasoftware.com/Seminars/TradeStation%20World%2005.pdf | ||
| 37 | # http://www.mesasoftware.com/Seminars/Seminars/TSWorld05.ppt | ||
| 38 | # (View the powerpoint with google.) | ||
| 39 | # Powerpoint summary by John Ehlers of several of his and other averages. | ||
| 40 | # View in google, | ||
| 41 | # * A Laguerre filter warps time in the filter coefficients | ||
| 42 | # - Enables extreme smoothing with just a few filter terms | ||
| 43 | # * A NonLinear Laguerre filter measures the difference between the | ||
| 44 | # current price and the last computed filter output. | ||
| 45 | # - Objective is to drive this "error" to zero | ||
| 46 | # - The "error", normalized to the error range over a selected period | ||
| 47 | # is the alpha of the Laguerre filter | ||
| 48 | # | ||
| 49 | |||
| 50 | |||
| 51 | sub longname { __('Adaptive Laguerre Filter') } | ||
| 52 | sub shortname { __('Adaptive Laguerre') } | ||
| 53 | sub manual { __p('manual-node','Adaptive Laguerre Filter') } | ||
| 54 | |||
| 55 | use constant | ||
| 56 | { type => 'average', | ||
| 57 | parameter_info => [ { name => __('Days'), | ||
| 58 | key => 'adaptive_laguerre_filter_days', | ||
| 59 | type => 'integer', | ||
| 60 | minimum => 1, | ||
| 61 | default => 20 } ], | ||
| 62 | }; | ||
| 63 | |||
| 64 | sub new { | ||
| 65 | my ($class, $parent, $N) = @_; | ||
| 66 | |||
| 67 | $N //= parameter_info()->[0]->{'default'}; | ||
| 68 | ($N >= 1) || croak "Adaptive Laguerre Filter bad N: $N"; | ||
| 69 | |||
| 70 | return $class->SUPER::new | ||
| 71 | (parent => $parent, | ||
| 72 | parameters => [ $N ], | ||
| 73 | arrays => { values => [] }, | ||
| 74 | array_aliases => { }); | ||
| 75 | } | ||
| 76 | |||
| 77 | sub proc { | ||
| 78 | my ($class, $N) = @_; | ||
| 79 | my $proc_laguerre_and_alpha = $class->proc_laguerre_and_alpha($N); | ||
| 80 | return sub { | ||
| 81 | return ($proc_laguerre_and_alpha->(@_))[0]; | ||
| 82 | }; | ||
| 83 | } | ||
| 84 | sub proc_laguerre_and_alpha { | ||
| 85 | my ($class, $N) = @_; | ||
| 86 | |||
| 87 | my $laguerre_proc | ||
| 88 | = App::Chart::Series::Derived::LaguerreFilter->proc_for_alpha(); | ||
| 89 | my $williams_proc = App::Chart::Series::Derived::WilliamsR->proc($N); | ||
| 90 | my $median_proc = App::Chart::Series::Derived::Median->proc(5); | ||
| 91 | my $alpha = 0.2; | ||
| 92 | my $prev; | ||
| 93 | |||
| 94 | return sub { | ||
| 95 | my ($value) = @_; | ||
| 96 | if (defined $prev) { | ||
| 97 | my $w = $williams_proc->(undef, undef, abs ($value - $prev)); | ||
| 98 | $alpha = $median_proc->(0.01 * ($w + 100)); # 0 to 1 | ||
| 99 | } | ||
| 100 | return (($prev = $laguerre_proc->($value, $alpha)), | ||
| 101 | $alpha); | ||
| 102 | }; | ||
| 103 | } | ||
| 104 | |||
| 105 | |||
| 106 | # warmup_count() gives a fixed amount, based on the worst-case EMA alphas | ||
| 107 | # all the slowest possible. It ends up being 1656 which is hugely more than | ||
| 108 | # needed in practice. | ||
| 109 | # | ||
| 110 | # warmup_count_for_position() calculates a value on actual data, working | ||
| 111 | # backwards. In practice it's as little as about 100. | ||
| 112 | # | ||
| 113 | sub warmup_count { | ||
| 114 | my ($self_or_class, $N) = @_; | ||
| 115 | |||
| 116 | # FIXME: this is a big over-estimate | ||
| 117 | return $N + App::Chart::Series::Derived::LaguerreFilter->warmup_count(0.01); | ||
| 118 | } | ||
| 119 | |||
| 120 | ### AdaptiveLaguerre warmup_count(): __PACKAGE__->warmup_count(parameter_info()->[0]->{'default'}) | ||
| 121 | |||
| 122 | 1; | ||
| 123 | __END__ |