SLOYD / Math-LOESS-0.0000_04 / README 

            

              
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138

              
NAME
    Math::LOESS - Perl wrapper of the Locally-Weighted Regression package
    originally written by Cleveland, et al.
VERSION
    version 0.0000_04
SYNOPSIS
        use Math::LOESS;
     
        my $loess = Math::LOESS->new(x => $x, y => $y);
     
        $loess->fit();
        my $fitted_values = $loess->outputs->fitted_values;
     
        print $loess->summary();
     
        my $prediction = $loess->predict($new_data, 1);
        my $confidence_intervals = $prediction->confidence(0.05);
        print $confidence_internals->{fit};
        print $confidence_internals->{upper};
        print $confidence_internals->{lower};
CONSTRUCTION
        new((Piddle1D|Piddle2D) :$x, Piddle1D :$y, Piddle1D :$weights=undef,
            Num :$span=0.75, Str :$family='gaussian')
    Arguments:
      * $x
      A ($n, $p) piddle for x data, where $p is number of predictors. It's
      possible to have at most 8 predictors.
      * $y
      A ($n, 1) piddle for y data.
      * $weights
      Optional ($n, 1) piddle for weights to be given to individual
      observations. By default, an unweighted fit is carried out (all the
      weights are one).
      * $span
      The parameter controls the degree of smoothing. Default is 0.75.
      For span < 1, the neighbourhood used for the fit includes proportion
      span of the points, and these have tricubic weighting (proportional
      to (1 - (dist/maxdist)^3)^3). For span > 1, all points are used, with
      the "maximum distance" assumed to be span^(1/p) times the actual
      maximum distance for p explanatory variables.
      When provided as a construction parameter, it is like a shortcut for,
          $loess->model->span($span);
      * $family
      If "gaussian" fitting is by least-squares, and if "symmetric" a
      re-descending M estimator is used with Tukey's biweight function.
      When provided as a construction parameter, it is like a shortcut for,
          $loess->model->family($family);
NAME
    Math::LOESS - Perl wrapper of the Locally-Weighted Regression package
    originally written by Cleveland, et al.
ATTRIBUTES
 model
    Get an Math::LOESS::Model object.
 outputs
    Get an Math::LOESS::Outputs object.
 x
    Get input x data as a piddle.
 y
    Get input y data as a piddle.
 weights
    Get input weights data as a piddle.
 activated
    Returns a true value if the object's fit() method has been called.
METHODS
 fit
        fit()
 predict
        predict((Piddle1D|Piddle2D) $newdata, Bool $stderr=false)
    Returns a Math::LOESS::Prediction object.
 summary
        summary()
    Returns a summary string. For example,
        print $loess->summary();
SEE ALSO
    https://en.wikipedia.org/wiki/Local_regression
    PDL
AUTHOR
    Stephan Loyd <sloyd@cpan.org>
COPYRIGHT AND LICENSE
    This software is copyright (c) 2019 by Stephan Loyd.
    This is free software; you can redistribute it and/or modify it under
    the same terms as the Perl 5 programming language system itself.