| 1 | # Copyright 2007, 2009, 2010 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::LaguerreFilter; | ||
| 18 | 1 | ▶ | use 5.010; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 19 | 1 | ▶ | use strict; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 20 | 1 | ▶ | use warnings; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 21 | 1 | ▶ | use Carp; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 22 | 1 | ▶ | use List::Util qw(min max); |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 23 | 1 | ▶ | use Locale::TextDomain ('App-Chart'); |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 24 | |||
| 25 | 1 | ▶ | use base 'App::Chart::Series::Indicator'; |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 26 | use App::Chart::Series::Calculation; | ||
| 27 | use App::Chart::Series::Derived::EMA; | ||
| 28 | use App::Chart::Series::Derived::EMAx2; | ||
| 29 | |||
| 30 | # http://www.mesasoftware.com/technicalpapers.htm | ||
| 31 | # http://www.mesasoftware.com/Papers/TIME%20WARP.pdf | ||
| 32 | # Paper by John Elhers. | ||
| 33 | # | ||
| 34 | # http://www.mesasoftware.com/seminars.htm | ||
| 35 | # http://www.mesasoftware.com/Seminars/TradeStation%20World%2005.pdf | ||
| 36 | # http://www.mesasoftware.com/Seminars/Seminars/TSWorld05.ppt | ||
| 37 | # (View the powerpoint with google.) | ||
| 38 | # Summary by John Ehlers of several of his and other averages. | ||
| 39 | # * A Laguerre filter warps time in the filter coefficients | ||
| 40 | # - Enables extreme smoothing with just a few filter terms | ||
| 41 | # * A NonLinear Laguerre filter measures the difference between the | ||
| 42 | # current price and the last computed filter output. | ||
| 43 | # - Objective is to drive this "error" to zero | ||
| 44 | # - The "error", normalized to the error range over a selected period | ||
| 45 | # is the alpha of the Laguerre filter | ||
| 46 | # | ||
| 47 | |||
| 48 | |||
| 49 | sub longname { __('Laguerre Filter') } | ||
| 50 | sub shortname { __('Laguerre') } | ||
| 51 | sub manual { __p('manual-node','Laguerre Filter') } | ||
| 52 | |||
| 53 | use constant | ||
| 54 | { type => 'average', | ||
| 55 | parameter_info => [ { name => __('Alpha'), | ||
| 56 | key => 'laguerre_filter_alpha', | ||
| 57 | type => 'float', | ||
| 58 | minimum => 0, | ||
| 59 | maximum => 1, | ||
| 60 | default => 0.2, | ||
| 61 | decimals => 2, | ||
| 62 | step => 0.1 } ], | ||
| 63 | }; | ||
| 64 | |||
| 65 | sub new { | ||
| 66 | my ($class, $parent, $alpha) = @_; | ||
| 67 | |||
| 68 | $alpha //= parameter_info()->[0]->{'default'}; | ||
| 69 | ($alpha >= 0 && $alpha <= 1.0) || croak "Laguerre Filter bad alpha: $alpha"; | ||
| 70 | |||
| 71 | return $class->SUPER::new | ||
| 72 | (parent => $parent, | ||
| 73 | parameters => [ $alpha ], | ||
| 74 | arrays => { values => [] }, | ||
| 75 | array_aliases => { }); | ||
| 76 | } | ||
| 77 | |||
| 78 | sub proc { | ||
| 79 | my ($class, $alpha) = @_; | ||
| 80 | $alpha = max (0.00001, min (0.99999, $alpha)); | ||
| 81 | my $proc_for_alpha = $class->proc_for_alpha(); | ||
| 82 | return sub { | ||
| 83 | my ($value) = @_; | ||
| 84 | return $proc_for_alpha->($value, $alpha); | ||
| 85 | }; | ||
| 86 | } | ||
| 87 | |||
| 88 | sub proc_for_alpha { | ||
| 89 | my ($class) = @_; | ||
| 90 | |||
| 91 | my $a_sum = 0; | ||
| 92 | my $a_div = 0; | ||
| 93 | my ($a_prev, $a_pdiv); | ||
| 94 | |||
| 95 | my $b_sum = 0; | ||
| 96 | my $b_div = 0; | ||
| 97 | my ($b_prev, $b_pdiv); | ||
| 98 | |||
| 99 | my $c_sum = 0; | ||
| 100 | my $c_div = 0; | ||
| 101 | my ($c_prev, $c_pdiv); | ||
| 102 | |||
| 103 | my $d_sum = 0; | ||
| 104 | my $d_div = 0; | ||
| 105 | |||
| 106 | return sub { | ||
| 107 | my ($value, $alpha) = @_; | ||
| 108 | my $f = 1 - $alpha; | ||
| 109 | |||
| 110 | $a_prev = $a_sum; | ||
| 111 | $a_pdiv = $a_div; | ||
| 112 | $a_sum = $value * $alpha + $a_sum * $f; | ||
| 113 | $a_div = $alpha + $a_div * $f; | ||
| 114 | |||
| 115 | $b_prev = $b_sum; | ||
| 116 | $b_pdiv = $b_div; | ||
| 117 | $b_sum = $a_prev + ($b_sum - $a_sum) * $f; | ||
| 118 | $b_div = $a_pdiv + ($b_div - $a_div) * $f; | ||
| 119 | |||
| 120 | $c_prev = $c_sum; | ||
| 121 | $c_pdiv = $c_div; | ||
| 122 | $c_sum = $b_prev + ($c_sum - $b_sum) * $f; | ||
| 123 | $c_div = $b_pdiv + ($c_div - $b_div) * $f; | ||
| 124 | |||
| 125 | $d_sum = $c_prev + ($d_sum - $c_sum) * $f; | ||
| 126 | $d_div = $c_pdiv + ($d_div - $c_div) * $f; | ||
| 127 | |||
| 128 | return ($a_sum/$a_div | ||
| 129 | + 2 * $b_sum/$b_div | ||
| 130 | + 2 * $c_sum/$c_div | ||
| 131 | + $d_sum/$d_div) / 6.0; | ||
| 132 | }; | ||
| 133 | } | ||
| 134 | |||
| 135 | # warmup_count() gives a fixed amount, based on the worst-case EMA alphas | ||
| 136 | # all the slowest possible. It ends up being 1656 which is hugely more than | ||
| 137 | # needed in practice. | ||
| 138 | # | ||
| 139 | # warmup_count_for_position() calculates a value on actual data, working | ||
| 140 | # backwards. In practice it's as little as about 100. | ||
| 141 | # | ||
| 142 | sub warmup_count { | ||
| 143 | my ($self_or_class, $alpha) = @_; | ||
| 144 | |||
| 145 | $alpha = max (0.00001, min (0.99999, $alpha)); | ||
| 146 | my $f = 1 - $alpha; | ||
| 147 | |||
| 148 | return App::Chart::Series::Derived::EMAx2::bsearch_first_true | ||
| 149 | (sub { | ||
| 150 | my ($i) = @_; | ||
| 151 | return (laguerre_omitted($f,$i) | ||
| 152 | <= App::Chart::Series::Derived::EMA::WARMUP_OMITTED_FRACTION); | ||
| 153 | }, | ||
| 154 | App::Chart::Series::Derived::EMA::alpha_to_N($alpha)); | ||
| 155 | } | ||
| 156 | |||
| 157 | # see devel/ema-omitted.pl | ||
| 158 | sub laguerre_omitted { | ||
| 159 | my ($f, $k) = @_; | ||
| 160 | return | ||
| 161 | $f ** ($k-2) | ||
| 162 | * ((1/36*$k**2 + -1/36)*$k | ||
| 163 | + ($f # f^($k-1) | ||
| 164 | * ((1/4*$k + 1/4)*$k | ||
| 165 | + ($f # f^$k | ||
| 166 | * ((((-1/12*$k + -1/6)*$k + 3/4)*$k + 5/6) | ||
| 167 | + ($f # f^($k+1) | ||
| 168 | * (((-1/2*$k + -1)*$k + 1/2) | ||
| 169 | + ($f # f^($k+2) | ||
| 170 | * ((((1/12*$k + 1/3)*$k + -5/12)*$k + -2/3) | ||
| 171 | + ($f # f^($k+3) | ||
| 172 | * (((1/4*$k + 3/4)*$k + 1/2) | ||
| 173 | + ($f # f^($k+4) | ||
| 174 | * ((((-1/36*$k + -1/6)*$k + -11/36)*$k + -1/6) | ||
| 175 | ))))))))))))); | ||
| 176 | } | ||
| 177 | |||
| 178 | 1; | ||
| 179 | __END__ | ||