1
|
|
|
# -*- coding: utf-8 -*- |
2
|
|
|
|
3
|
|
|
# Copyright 2014-2018 by Christopher C. Little. |
4
|
|
|
# This file is part of Abydos. |
5
|
|
|
# |
6
|
|
|
# Abydos is free software: you can redistribute it and/or modify |
7
|
|
|
# it under the terms of the GNU General Public License as published by |
8
|
|
|
# the Free Software Foundation, either version 3 of the License, or |
9
|
|
|
# (at your option) any later version. |
10
|
|
|
# |
11
|
|
|
# Abydos is distributed in the hope that it will be useful, |
12
|
|
|
# but WITHOUT ANY WARRANTY; without even the implied warranty of |
13
|
|
|
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
14
|
|
|
# GNU General Public License for more details. |
15
|
|
|
# |
16
|
|
|
# You should have received a copy of the GNU General Public License |
17
|
|
|
# along with Abydos. If not, see <http://www.gnu.org/licenses/>. |
18
|
|
|
|
19
|
1 |
|
"""abydos.distance._monge_elkan. |
20
|
|
|
|
21
|
|
|
Monge-Elkan similarity & distance |
22
|
|
|
""" |
23
|
|
|
|
24
|
1 |
|
from __future__ import ( |
25
|
|
|
absolute_import, |
26
|
|
|
division, |
27
|
|
|
print_function, |
28
|
|
|
unicode_literals, |
29
|
|
|
) |
30
|
|
|
|
31
|
1 |
|
from ._distance import _Distance |
32
|
1 |
|
from ._levenshtein import sim_levenshtein |
33
|
1 |
|
from ..tokenizer import QGrams |
34
|
|
|
|
35
|
1 |
|
__all__ = ['MongeElkan', 'dist_monge_elkan', 'sim_monge_elkan'] |
36
|
|
|
|
37
|
|
|
|
38
|
1 |
|
class MongeElkan(_Distance): |
39
|
|
|
"""Monge-Elkan similarity. |
40
|
|
|
|
41
|
|
|
Monge-Elkan is defined in :cite:`Monge:1996`. |
42
|
|
|
|
43
|
|
|
Note: Monge-Elkan is NOT a symmetric similarity algorithm. Thus, the |
44
|
|
|
similarity of src to tar is not necessarily equal to the similarity of |
45
|
|
|
tar to src. If the symmetric argument is True, a symmetric value is |
46
|
|
|
calculated, at the cost of doubling the computation time (since |
47
|
|
|
:math:`sim_{Monge-Elkan}(src, tar)` and :math:`sim_{Monge-Elkan}(tar, src)` |
48
|
|
|
are both calculated and then averaged). |
49
|
|
|
""" |
50
|
|
|
|
51
|
1 |
|
def sim(self, src, tar, sim_func=sim_levenshtein, symmetric=False): |
52
|
|
|
"""Return the Monge-Elkan similarity of two strings. |
53
|
|
|
|
54
|
|
|
Parameters |
55
|
|
|
---------- |
56
|
|
|
src : str |
57
|
|
|
Source string for comparison |
58
|
|
|
tar : str |
59
|
|
|
Target string for comparison |
60
|
|
|
sim_func : function |
61
|
|
|
The internal similarity metric to employ |
62
|
|
|
symmetric : bool |
63
|
|
|
Return a symmetric similarity measure |
64
|
|
|
|
65
|
|
|
Returns |
66
|
|
|
------- |
67
|
|
|
float |
68
|
|
|
Monge-Elkan similarity |
69
|
|
|
|
70
|
|
|
Examples |
71
|
|
|
-------- |
72
|
|
|
>>> cmp = MongeElkan() |
73
|
|
|
>>> cmp.sim('cat', 'hat') |
74
|
|
|
0.75 |
75
|
|
|
>>> round(cmp.sim('Niall', 'Neil'), 12) |
76
|
|
|
0.666666666667 |
77
|
|
|
>>> round(cmp.sim('aluminum', 'Catalan'), 12) |
78
|
|
|
0.388888888889 |
79
|
|
|
>>> cmp.sim('ATCG', 'TAGC') |
80
|
|
|
0.5 |
81
|
|
|
|
82
|
|
|
""" |
83
|
1 |
|
if src == tar: |
84
|
1 |
|
return 1.0 |
85
|
|
|
|
86
|
1 |
|
q_src = sorted(QGrams(src).elements()) |
87
|
1 |
|
q_tar = sorted(QGrams(tar).elements()) |
88
|
|
|
|
89
|
1 |
|
if not q_src or not q_tar: |
90
|
1 |
|
return 0.0 |
91
|
|
|
|
92
|
1 |
|
sum_of_maxes = 0 |
93
|
1 |
|
for q_s in q_src: |
94
|
1 |
|
max_sim = float('-inf') |
95
|
1 |
|
for q_t in q_tar: |
96
|
1 |
|
max_sim = max(max_sim, sim_func(q_s, q_t)) |
97
|
1 |
|
sum_of_maxes += max_sim |
98
|
1 |
|
sim_em = sum_of_maxes / len(q_src) |
99
|
|
|
|
100
|
1 |
|
if symmetric: |
101
|
1 |
|
sim_em = (sim_em + self.sim(tar, src, sim_func, False)) / 2 |
102
|
|
|
|
103
|
1 |
|
return sim_em |
104
|
|
|
|
105
|
|
|
|
106
|
1 |
|
def sim_monge_elkan(src, tar, sim_func=sim_levenshtein, symmetric=False): |
107
|
|
|
"""Return the Monge-Elkan similarity of two strings. |
108
|
|
|
|
109
|
|
|
This is a wrapper for :py:meth:`MongeElkan.sim`. |
110
|
|
|
|
111
|
|
|
Parameters |
112
|
|
|
---------- |
113
|
|
|
src : str |
114
|
|
|
Source string for comparison |
115
|
|
|
tar : str |
116
|
|
|
Target string for comparison |
117
|
|
|
sim_func : function |
118
|
|
|
Rhe internal similarity metric to employ |
119
|
|
|
symmetric : bool |
120
|
|
|
Return a symmetric similarity measure |
121
|
|
|
|
122
|
|
|
Returns |
123
|
|
|
------- |
124
|
|
|
float |
125
|
|
|
Monge-Elkan similarity |
126
|
|
|
|
127
|
|
|
Examples |
128
|
|
|
-------- |
129
|
|
|
>>> sim_monge_elkan('cat', 'hat') |
130
|
|
|
0.75 |
131
|
|
|
>>> round(sim_monge_elkan('Niall', 'Neil'), 12) |
132
|
|
|
0.666666666667 |
133
|
|
|
>>> round(sim_monge_elkan('aluminum', 'Catalan'), 12) |
134
|
|
|
0.388888888889 |
135
|
|
|
>>> sim_monge_elkan('ATCG', 'TAGC') |
136
|
|
|
0.5 |
137
|
|
|
|
138
|
|
|
""" |
139
|
1 |
|
return MongeElkan().sim(src, tar, sim_func, symmetric) |
140
|
|
|
|
141
|
|
|
|
142
|
1 |
|
def dist_monge_elkan(src, tar, sim_func=sim_levenshtein, symmetric=False): |
143
|
|
|
"""Return the Monge-Elkan distance between two strings. |
144
|
|
|
|
145
|
|
|
This is a wrapper for :py:meth:`MongeElkan.dist`. |
146
|
|
|
|
147
|
|
|
Parameters |
148
|
|
|
---------- |
149
|
|
|
src : str |
150
|
|
|
Source string for comparison |
151
|
|
|
tar : str |
152
|
|
|
Target string for comparison |
153
|
|
|
sim_func : function |
154
|
|
|
The internal similarity metric to employ |
155
|
|
|
symmetric : bool |
156
|
|
|
Return a symmetric similarity measure |
157
|
|
|
|
158
|
|
|
Returns |
159
|
|
|
------- |
160
|
|
|
float |
161
|
|
|
Monge-Elkan distance |
162
|
|
|
|
163
|
|
|
Examples |
164
|
|
|
-------- |
165
|
|
|
>>> dist_monge_elkan('cat', 'hat') |
166
|
|
|
0.25 |
167
|
|
|
>>> round(dist_monge_elkan('Niall', 'Neil'), 12) |
168
|
|
|
0.333333333333 |
169
|
|
|
>>> round(dist_monge_elkan('aluminum', 'Catalan'), 12) |
170
|
|
|
0.611111111111 |
171
|
|
|
>>> dist_monge_elkan('ATCG', 'TAGC') |
172
|
|
|
0.5 |
173
|
|
|
|
174
|
|
|
""" |
175
|
1 |
|
return MongeElkan().dist(src, tar, sim_func, symmetric) |
176
|
|
|
|
177
|
|
|
|
178
|
|
|
if __name__ == '__main__': |
179
|
|
|
import doctest |
180
|
|
|
|
181
|
|
|
doctest.testmod() |
182
|
|
|
|