forked from antononcube/MathematicaForPrediction
-
Notifications
You must be signed in to change notification settings - Fork 0
/
OutlierIdentifiers.m
217 lines (145 loc) · 8 KB
/
OutlierIdentifiers.m
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
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
(*
Implementation of one dimensional outlier identifying algorithms in Mathematica
Copyright (C) 2013 Anton Antonov
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
Written by Anton Antonov,
antononcube @ gmail . com,
Windermere, Florida, USA.
*)
(*
Mathematica is (C) Copyright 1988-2019 Wolfram Research, Inc.
Protected by copyright law and international treaties.
Unauthorized reproduction or distribution subject to severe civil
and criminal penalties.
Mathematica is a registered trademark of Wolfram Research, Inc.
*)
(* Version 1.0 *)
(*
# In brief
This package contains definitions for detection and visualization of outliers in a list of numbers.
The purpose of the outlier detection algorithms is to find those elements in a list of numbers
that have values significantly higher or lower than the rest of the values.
Taking a certain number of elements with the highest values is not the same as an outlier detection,
but it can be used as a replacement.
# Usage
Let us consider the following set of 50 numbers:
SeedRandom[343]
pnts = RandomVariate[GammaDistribution[5, 1], 50]
Here we find the outliers using the HampelIdentifierParameters function:
OutlierIdentify[pnts, HampelIdentifierParameters]
Here we find the top outliers only:
OutlierIdentify[pnts, TopOutliers @* HampelIdentifierParameters]
(* {7.68192, 8.47235, <<9>>, 6.57855, 6.96975} *)
Here we find the top outliers positions:
OutlierPosition[pnts, TopOutliers @* HampelIdentifierParameters]
(* {3, 4, 6, 8, 13, 26, 34, 37, 38, 41, 42, 47, 48} *)
Here is the application of all outlier parameter finding functions in this package:
Through[ {HampelIdentifierParameters, SPLUSQuartileIdentifierParameters, QuartileIdentifierParameters}[pnts] ]
(* {{2.17496, 6.54877}, {-2.09104, 11.7803}, {0.572922, 7.50859}} *)
# References
[1] Ronald K. Pearson, “Mining Imperfect Data: Dealing with Contamination and Incomplete Records”, 2005, SIAM.
*)
BeginPackage["OutlierIdentifiers`"];
HampelIdentifierParameters::usage = "Returns Hampel outlier identifier parameters {L,U} for a list of numbers.";
QuartileIdentifierParameters::usage = "Returns quartile outlier identifier parameters {L,U} for a list of numbers.";
SPLUSQuartileIdentifierParameters::usage = "Returns SPLUS quartile outlier identifier parameters {L,U} for a list of numbers.";
OutlierIdentify::usage = "OutlierIdentify[data : {_?NumberQ...} | Association[ (_ -> _?NumberQ) ..], pars] \
applies outlier identifier parameters pars to a list of numbers dataArg.";
OutlierIdentifier::usage = "Synonym of OutlierIdentify.";
OutlierIdentifyLess::usage = "OutlierIdentifyLess[ data : {_?NumberQ...} | Association[ (_ -> _?NumberQ) ..], pars] \
applies outlier identifier parameters pars to a list of numbers data and takes the outliers with smallest values.";
TopOutliers::usage = "Changes the parameters {L,U} of an outlier identifier to {-Infinity,U}.";
BottomOutliers::usage = "Changes the parameters {L,U} of an outlier identifier to {L,Infinity}.";
HampelIdentifier::usage = "Shortcut for OutlierIdentify[#, HampelIdentifierParameters]& .";
QuartileIdentifier::usage = "Shortcut for OutlierIdentify[#, QuartileIdentifierParameters]& .";
SPLUSQuartileIdentifier::usage = "Shortcut for OutlierIdentify[#, SPLUSQuartileIdentifierParameters]& .";
OutlierPosition::usage = "OutlierPosition[ data : {_?NumberQ...}, pars] gives the positions of the outliers \
in data using the outlier identifier parameters pars. Top and bottom outliers can be found with
TopOutliers @* pars and BottomOutliers @* pars respectively.";
ListPlotOutliers::usage = "Plots a list of numbers and its outliers using ListPlot.";
ColorPlotOutliers::usage = "ColorPlotOutliers[oid___] makes a function for coloring the outliers in list point plots.";
Begin["`Private`"];
Clear[HampelIdentifierParameters];
HampelIdentifierParameters[data : {_?NumberQ...}] :=
Block[{x0 = Median[data], md},
md = 1.4826 * Median[Abs[data - x0]];
{x0 - md, x0 + md}
];
Clear[QuartileIdentifierParameters];
QuartileIdentifierParameters[data : {_?NumberQ...}] :=
Block[{xL, xU, x0},
{xL, x0, xU} = Quantile[data, {1 / 4, 1 / 2, 3 / 4}];
{x0 - (xU - xL), x0 + (xU - xL)}
];
Clear[SPLUSQuartileIdentifierParameters];
SPLUSQuartileIdentifierParameters[data : {_?NumberQ...}] :=
Block[{xL, xU},
If[Length[data] <= 4, Return[{Min[data], Max[data]}]];
{xL, xU} = Quantile[data, {1 / 4, 3 / 4}];
{xL - 1.5(xU - xL), xU + 1.5(xU - xL)}
];
Clear[TopOutliers, BottomOutliers];
TopOutliers[{xL_, xU_}] := {-Infinity, xU};
BottomOutliers[{xL_, xU_}] := {xL, Infinity};
(***********::Section:: ***********)
(* Identifiers *)
(**********************************)
Clear[OutlierIdentify, OutlierIdentifyLess];
OutlierIdentify[data : {_?NumberQ...}, outlierIdentifierParameters_ : HampelIdentifierParameters ] :=
Block[{xL, xU},
{xL, xU} = outlierIdentifierParameters[data];
Select[data, # < xL || xU < #&]
];
OutlierIdentify[ data : Association[ (_ -> _?NumberQ) ..], outlierIdentifierParameters_ : HampelIdentifierParameters ] :=
KeyTake[ data, OutlierPosition[data, outlierIdentifierParameters] ];
OutlierIdentifyLess[data : {_?NumberQ...} | Association[ (_ -> _?NumberQ) ..], outlierIdentifierParameters_ : HampelIdentifierParameters ] :=
OutlierIdentify[ data, BottomOutliers @* outlierIdentifierParameters ];
Clear[OutlierIdentifier];
OutlierIdentifier = OutlierIdentify;
Clear[HampelIdentifier];
HampelIdentifier[data_] := OutlierIdentify[data, HampelIdentifierParameters];
Clear[QuartileIdentifier];
QuartileIdentifier[data_] := OutlierIdentify[data, QuartileIdentifierParameters];
Clear[SPLUSQuartileIdentifier];
SPLUSQuartileIdentifier[data_] := OutlierIdentify[data, SPLUSQuartileIdentifierParameters];
Clear[OutlierPosition];
OutlierPosition[data : {_?NumberQ...}, outlierIdentifier_ : HampelIdentifierParameters] :=
Block[{cls, t},
cls = OutlierIdentify[data, outlierIdentifier];
t = Select[Transpose[{data, Range[Length[data]]}], MemberQ[cls, #[[1]]]&];
If[t === {}, {}, t[[All, 2]]]
];
OutlierPosition[data : Association[ (_ -> _?NumberQ) ... ], outlierIdentifier_ : HampelIdentifierParameters ] :=
Block[{pos},
pos = OutlierPosition[ Values[data], outlierIdentifier];
If[ pos === {}, {}, Keys[data][[pos]] ]
];
(*********** ::Section:: ***********)
(* Plot definitions *)
(***********************************)
Clear[ListPlotOutliers];
Options[ListPlotOutliers] = {PlotStyle -> {PointSize[0.015]}, PlotRange -> All, ImageSize -> 300};
ListPlotOutliers[ds_, outlierParameters_, optsArg___] :=
Block[{outliers, opts = optsArg, positionedOutliers},
If[!OptionQ[{opts}], opts = Options[ListPlotOutliers]];
outliers = OutlierIdentify[ds, outlierParameters];
If[outliers === {},
ListPlot[Transpose[{Range[Length[ds]], ds}], opts],
positionedOutliers = Select[Transpose[{Range[Length[ds]], ds}], MemberQ[outliers, #[[2]]]&];
ListPlot[{Transpose[{Range[Length[ds]], ds}], positionedOutliers}, opts]
]
];
ClearAll[ColorPlotOutliers];
ColorPlotOutliers[] := # /. {Point[ps_] :> {Point[ps], Red, Point[ps[[OutlierPosition[ps[[All, 2]]]]]]}} &;
ColorPlotOutliers[oid_] := # /. {Point[ps_] :> {Point[ps], Red, Point[ps[[OutlierPosition[ps[[All, 2]], oid]]]]}} &;
End[];
EndPackage[];