|
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
|
|
|
"""abydos.distance.compression. |
|
20
|
|
|
|
|
21
|
|
|
The distance.compression module implements compression distance measures. |
|
22
|
|
|
""" |
|
23
|
|
|
|
|
24
|
|
|
from __future__ import division, unicode_literals |
|
25
|
|
|
|
|
26
|
|
|
from codecs import encode |
|
27
|
|
|
from sys import modules |
|
28
|
|
|
|
|
29
|
|
|
from ..compression import arithmetic, rle |
|
30
|
|
|
|
|
31
|
|
|
try: |
|
32
|
|
|
import lzma |
|
33
|
|
|
except ImportError: # pragma: no cover |
|
34
|
|
|
# If the system lacks the lzma library, that's fine, but lzma compression |
|
35
|
|
|
# similarity won't be supported. |
|
36
|
|
|
lzma = None |
|
37
|
|
|
|
|
38
|
|
|
__all__ = ['dist_ncd_arith', 'dist_ncd_bwtrle', 'dist_ncd_bz2', |
|
39
|
|
|
'dist_ncd_lzma', 'dist_ncd_rle', 'dist_ncd_zlib', 'sim_ncd_arith', |
|
40
|
|
|
'sim_ncd_bwtrle', 'sim_ncd_bz2', 'sim_ncd_lzma', 'sim_ncd_rle', |
|
41
|
|
|
'sim_ncd_zlib'] |
|
42
|
|
|
|
|
43
|
|
|
|
|
44
|
|
|
def dist_ncd_arith(src, tar, probs=None): |
|
45
|
|
|
"""Return the NCD between two strings using arithmetic coding. |
|
46
|
|
|
|
|
47
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
48
|
|
|
|
|
49
|
|
|
:param str src: source string for comparison |
|
50
|
|
|
:param str tar: target string for comparison |
|
51
|
|
|
:param dict probs: a dictionary trained with arithmetic.train (for the |
|
52
|
|
|
arith compressor only) |
|
53
|
|
|
:returns: compression distance |
|
54
|
|
|
:rtype: float |
|
55
|
|
|
|
|
56
|
|
|
>>> dist_ncd_arith('cat', 'hat') |
|
57
|
|
|
0.5454545454545454 |
|
58
|
|
|
>>> dist_ncd_arith('Niall', 'Neil') |
|
59
|
|
|
0.6875 |
|
60
|
|
|
>>> dist_ncd_arith('aluminum', 'Catalan') |
|
61
|
|
|
0.8275862068965517 |
|
62
|
|
|
>>> dist_ncd_arith('ATCG', 'TAGC') |
|
63
|
|
|
0.6923076923076923 |
|
64
|
|
|
""" |
|
65
|
|
|
if src == tar: |
|
66
|
|
|
return 0.0 |
|
67
|
|
|
|
|
68
|
|
|
if probs is None: |
|
69
|
|
|
# lacking a reasonable dictionary, train on the strings themselves |
|
70
|
|
|
probs = arithmetic.train(src + tar) |
|
71
|
|
|
src_comp = arithmetic.encode(src, probs)[1] |
|
72
|
|
|
tar_comp = arithmetic.encode(tar, probs)[1] |
|
73
|
|
|
concat_comp = arithmetic.encode(src + tar, probs)[1] |
|
74
|
|
|
concat_comp2 = arithmetic.encode(tar + src, probs)[1] |
|
75
|
|
|
|
|
76
|
|
|
return ((min(concat_comp, concat_comp2) - min(src_comp, tar_comp)) / |
|
77
|
|
|
max(src_comp, tar_comp)) |
|
78
|
|
|
|
|
79
|
|
|
|
|
80
|
|
|
def sim_ncd_arith(src, tar, probs=None): |
|
81
|
|
|
"""Return the NCD similarity between two strings using arithmetic coding. |
|
82
|
|
|
|
|
83
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
84
|
|
|
|
|
85
|
|
|
:param str src: source string for comparison |
|
86
|
|
|
:param str tar: target string for comparison |
|
87
|
|
|
:param dict probs: a dictionary trained with arithmetic.train (for the |
|
88
|
|
|
arith compressor only) |
|
89
|
|
|
:returns: compression similarity |
|
90
|
|
|
:rtype: float |
|
91
|
|
|
|
|
92
|
|
|
>>> sim_ncd_arith('cat', 'hat') |
|
93
|
|
|
0.4545454545454546 |
|
94
|
|
|
>>> sim_ncd_arith('Niall', 'Neil') |
|
95
|
|
|
0.3125 |
|
96
|
|
|
>>> sim_ncd_arith('aluminum', 'Catalan') |
|
97
|
|
|
0.1724137931034483 |
|
98
|
|
|
>>> sim_ncd_arith('ATCG', 'TAGC') |
|
99
|
|
|
0.3076923076923077 |
|
100
|
|
|
""" |
|
101
|
|
|
return 1-dist_ncd_arith(src, tar, probs) |
|
102
|
|
|
|
|
103
|
|
|
|
|
104
|
|
|
def dist_ncd_rle(src, tar, use_bwt=False): |
|
105
|
|
|
"""Return the NCD between two strings using RLE. |
|
106
|
|
|
|
|
107
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
108
|
|
|
|
|
109
|
|
|
:param str src: source string for comparison |
|
110
|
|
|
:param str tar: target string for comparison |
|
111
|
|
|
:param bool use_bwt: boolean indicating whether to perform BWT encoding |
|
112
|
|
|
before RLE encoding |
|
113
|
|
|
:returns: compression distance |
|
114
|
|
|
:rtype: float |
|
115
|
|
|
|
|
116
|
|
|
>>> dist_ncd_rle('cat', 'hat') |
|
117
|
|
|
1.0 |
|
118
|
|
|
>>> dist_ncd_rle('Niall', 'Neil') |
|
119
|
|
|
1.0 |
|
120
|
|
|
>>> dist_ncd_rle('aluminum', 'Catalan') |
|
121
|
|
|
1.0 |
|
122
|
|
|
>>> dist_ncd_rle('ATCG', 'TAGC') |
|
123
|
|
|
1.0 |
|
124
|
|
|
""" |
|
125
|
|
|
if src == tar: |
|
126
|
|
|
return 0.0 |
|
127
|
|
|
|
|
128
|
|
|
src_comp = rle.encode(src, use_bwt) |
|
129
|
|
|
tar_comp = rle.encode(tar, use_bwt) |
|
130
|
|
|
concat_comp = rle.encode(src + tar, use_bwt) |
|
131
|
|
|
concat_comp2 = rle.encode(tar + src, use_bwt) |
|
132
|
|
|
|
|
133
|
|
|
return ((min(len(concat_comp), len(concat_comp2)) - |
|
134
|
|
|
min(len(src_comp), len(tar_comp))) / |
|
135
|
|
|
max(len(src_comp), len(tar_comp))) |
|
136
|
|
|
|
|
137
|
|
|
|
|
138
|
|
|
def sim_ncd_rle(src, tar, use_bwt=False): |
|
139
|
|
|
"""Return the NCD similarity between two strings using RLE. |
|
140
|
|
|
|
|
141
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
142
|
|
|
|
|
143
|
|
|
:param str src: source string for comparison |
|
144
|
|
|
:param str tar: target string for comparison |
|
145
|
|
|
:param bool use_bwt: boolean indicating whether to perform BWT encoding |
|
146
|
|
|
before RLE encoding |
|
147
|
|
|
:returns: compression similarity |
|
148
|
|
|
:rtype: float |
|
149
|
|
|
|
|
150
|
|
|
>>> sim_ncd_rle('cat', 'hat') |
|
151
|
|
|
0.0 |
|
152
|
|
|
>>> sim_ncd_rle('Niall', 'Neil') |
|
153
|
|
|
0.0 |
|
154
|
|
|
>>> sim_ncd_rle('aluminum', 'Catalan') |
|
155
|
|
|
0.0 |
|
156
|
|
|
>>> sim_ncd_rle('ATCG', 'TAGC') |
|
157
|
|
|
0.0 |
|
158
|
|
|
""" |
|
159
|
|
|
return 1 - dist_ncd_rle(src, tar, use_bwt) |
|
160
|
|
|
|
|
161
|
|
|
|
|
162
|
|
|
def dist_ncd_bwtrle(src, tar): |
|
163
|
|
|
"""Return the NCD between two strings using BWT plus RLE. |
|
164
|
|
|
|
|
165
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
166
|
|
|
|
|
167
|
|
|
:param str src: source string for comparison |
|
168
|
|
|
:param str tar: target string for comparison |
|
169
|
|
|
:returns: compression distance |
|
170
|
|
|
:rtype: float |
|
171
|
|
|
|
|
172
|
|
|
>>> dist_ncd_bwtrle('cat', 'hat') |
|
173
|
|
|
0.75 |
|
174
|
|
|
>>> dist_ncd_bwtrle('Niall', 'Neil') |
|
175
|
|
|
0.8333333333333334 |
|
176
|
|
|
>>> dist_ncd_bwtrle('aluminum', 'Catalan') |
|
177
|
|
|
1.0 |
|
178
|
|
|
>>> dist_ncd_bwtrle('ATCG', 'TAGC') |
|
179
|
|
|
0.8 |
|
180
|
|
|
""" |
|
181
|
|
|
return dist_ncd_rle(src, tar, True) |
|
182
|
|
|
|
|
183
|
|
|
|
|
184
|
|
|
def sim_ncd_bwtrle(src, tar): |
|
185
|
|
|
"""Return the NCD similarity between two strings using BWT plus RLE. |
|
186
|
|
|
|
|
187
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
188
|
|
|
|
|
189
|
|
|
:param str src: source string for comparison |
|
190
|
|
|
:param str tar: target string for comparison |
|
191
|
|
|
:returns: compression similarity |
|
192
|
|
|
:rtype: float |
|
193
|
|
|
|
|
194
|
|
|
>>> sim_ncd_bwtrle('cat', 'hat') |
|
195
|
|
|
0.25 |
|
196
|
|
|
>>> sim_ncd_bwtrle('Niall', 'Neil') |
|
197
|
|
|
0.16666666666666663 |
|
198
|
|
|
>>> sim_ncd_bwtrle('aluminum', 'Catalan') |
|
199
|
|
|
0.0 |
|
200
|
|
|
>>> sim_ncd_bwtrle('ATCG', 'TAGC') |
|
201
|
|
|
0.19999999999999996 |
|
202
|
|
|
""" |
|
203
|
|
|
return 1 - dist_ncd_bwtrle(src, tar) |
|
204
|
|
|
|
|
205
|
|
|
|
|
206
|
|
View Code Duplication |
def dist_ncd_zlib(src, tar): |
|
|
|
|
|
|
207
|
|
|
"""Return the NCD between two strings using zlib compression. |
|
208
|
|
|
|
|
209
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
210
|
|
|
|
|
211
|
|
|
:param str src: source string for comparison |
|
212
|
|
|
:param str tar: target string for comparison |
|
213
|
|
|
:returns: compression distance |
|
214
|
|
|
:rtype: float |
|
215
|
|
|
|
|
216
|
|
|
>>> dist_ncd_zlib('cat', 'hat') |
|
217
|
|
|
0.3333333333333333 |
|
218
|
|
|
>>> dist_ncd_zlib('Niall', 'Neil') |
|
219
|
|
|
0.45454545454545453 |
|
220
|
|
|
>>> dist_ncd_zlib('aluminum', 'Catalan') |
|
221
|
|
|
0.5714285714285714 |
|
222
|
|
|
>>> dist_ncd_zlib('ATCG', 'TAGC') |
|
223
|
|
|
0.4 |
|
224
|
|
|
""" |
|
225
|
|
|
if src == tar: |
|
226
|
|
|
return 0.0 |
|
227
|
|
|
|
|
228
|
|
|
src = src.encode('utf-8') |
|
229
|
|
|
tar = tar.encode('utf-8') |
|
230
|
|
|
|
|
231
|
|
|
src_comp = encode(src, 'zlib_codec')[2:] |
|
232
|
|
|
tar_comp = encode(tar, 'zlib_codec')[2:] |
|
233
|
|
|
concat_comp = encode(src + tar, 'zlib_codec')[2:] |
|
234
|
|
|
concat_comp2 = encode(tar + src, 'zlib_codec')[2:] |
|
235
|
|
|
|
|
236
|
|
|
return ((min(len(concat_comp), len(concat_comp2)) - |
|
237
|
|
|
min(len(src_comp), len(tar_comp))) / |
|
238
|
|
|
max(len(src_comp), len(tar_comp))) |
|
239
|
|
|
|
|
240
|
|
|
|
|
241
|
|
|
def sim_ncd_zlib(src, tar): |
|
242
|
|
|
"""Return the NCD similarity between two strings using zlib compression. |
|
243
|
|
|
|
|
244
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
245
|
|
|
|
|
246
|
|
|
:param str src: source string for comparison |
|
247
|
|
|
:param str tar: target string for comparison |
|
248
|
|
|
:returns: compression similarity |
|
249
|
|
|
:rtype: float |
|
250
|
|
|
|
|
251
|
|
|
>>> sim_ncd_zlib('cat', 'hat') |
|
252
|
|
|
0.6666666666666667 |
|
253
|
|
|
>>> sim_ncd_zlib('Niall', 'Neil') |
|
254
|
|
|
0.5454545454545454 |
|
255
|
|
|
>>> sim_ncd_zlib('aluminum', 'Catalan') |
|
256
|
|
|
0.4285714285714286 |
|
257
|
|
|
>>> sim_ncd_zlib('ATCG', 'TAGC') |
|
258
|
|
|
0.6 |
|
259
|
|
|
""" |
|
260
|
|
|
return 1 - dist_ncd_zlib(src, tar) |
|
261
|
|
|
|
|
262
|
|
|
|
|
263
|
|
View Code Duplication |
def dist_ncd_bz2(src, tar): |
|
|
|
|
|
|
264
|
|
|
"""Return the NCD between two strings using bz2 compression. |
|
265
|
|
|
|
|
266
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
267
|
|
|
|
|
268
|
|
|
:param str src: source string for comparison |
|
269
|
|
|
:param str tar: target string for comparison |
|
270
|
|
|
:returns: compression distance |
|
271
|
|
|
:rtype: float |
|
272
|
|
|
|
|
273
|
|
|
>>> dist_ncd_bz2('cat', 'hat') |
|
274
|
|
|
0.08 |
|
275
|
|
|
>>> dist_ncd_bz2('Niall', 'Neil') |
|
276
|
|
|
0.037037037037037035 |
|
277
|
|
|
>>> dist_ncd_bz2('aluminum', 'Catalan') |
|
278
|
|
|
0.20689655172413793 |
|
279
|
|
|
>>> dist_ncd_bz2('ATCG', 'TAGC') |
|
280
|
|
|
0.037037037037037035 |
|
281
|
|
|
""" |
|
282
|
|
|
if src == tar: |
|
283
|
|
|
return 0.0 |
|
284
|
|
|
|
|
285
|
|
|
src = src.encode('utf-8') |
|
286
|
|
|
tar = tar.encode('utf-8') |
|
287
|
|
|
|
|
288
|
|
|
src_comp = encode(src, 'bz2_codec')[15:] |
|
289
|
|
|
tar_comp = encode(tar, 'bz2_codec')[15:] |
|
290
|
|
|
concat_comp = encode(src + tar, 'bz2_codec')[15:] |
|
291
|
|
|
concat_comp2 = encode(tar + src, 'bz2_codec')[15:] |
|
292
|
|
|
|
|
293
|
|
|
return ((min(len(concat_comp), len(concat_comp2)) - |
|
294
|
|
|
min(len(src_comp), len(tar_comp))) / |
|
295
|
|
|
max(len(src_comp), len(tar_comp))) |
|
296
|
|
|
|
|
297
|
|
|
|
|
298
|
|
|
def sim_ncd_bz2(src, tar): |
|
299
|
|
|
"""Return the NCD similarity between two strings using bz2 compression. |
|
300
|
|
|
|
|
301
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
302
|
|
|
|
|
303
|
|
|
:param str src: source string for comparison |
|
304
|
|
|
:param str tar: target string for comparison |
|
305
|
|
|
:returns: compression similarity |
|
306
|
|
|
:rtype: float |
|
307
|
|
|
|
|
308
|
|
|
>>> sim_ncd_bz2('cat', 'hat') |
|
309
|
|
|
0.92 |
|
310
|
|
|
>>> sim_ncd_bz2('Niall', 'Neil') |
|
311
|
|
|
0.962962962962963 |
|
312
|
|
|
>>> sim_ncd_bz2('aluminum', 'Catalan') |
|
313
|
|
|
0.7931034482758621 |
|
314
|
|
|
>>> sim_ncd_bz2('ATCG', 'TAGC') |
|
315
|
|
|
0.962962962962963 |
|
316
|
|
|
""" |
|
317
|
|
|
return 1 - dist_ncd_bz2(src, tar) |
|
318
|
|
|
|
|
319
|
|
|
|
|
320
|
|
|
def dist_ncd_lzma(src, tar): |
|
321
|
|
|
"""Return the NCD between two strings using lzma compression. |
|
322
|
|
|
|
|
323
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
324
|
|
|
|
|
325
|
|
|
:param str src: source string for comparison |
|
326
|
|
|
:param str tar: target string for comparison |
|
327
|
|
|
:returns: compression distance |
|
328
|
|
|
:rtype: float |
|
329
|
|
|
|
|
330
|
|
|
>>> dist_ncd_lzma('cat', 'hat') |
|
331
|
|
|
0.08695652173913043 |
|
332
|
|
|
>>> dist_ncd_lzma('Niall', 'Neil') |
|
333
|
|
|
0.16 |
|
334
|
|
|
>>> dist_ncd_lzma('aluminum', 'Catalan') |
|
335
|
|
|
0.16 |
|
336
|
|
|
>>> dist_ncd_lzma('ATCG', 'TAGC') |
|
337
|
|
|
0.08695652173913043 |
|
338
|
|
|
""" |
|
339
|
|
|
if src == tar: |
|
340
|
|
|
return 0.0 |
|
341
|
|
|
|
|
342
|
|
|
src = src.encode('utf-8') |
|
343
|
|
|
tar = tar.encode('utf-8') |
|
344
|
|
|
|
|
345
|
|
|
if 'lzma' in modules: |
|
346
|
|
|
src_comp = lzma.compress(src)[14:] |
|
347
|
|
|
tar_comp = lzma.compress(tar)[14:] |
|
348
|
|
|
concat_comp = lzma.compress(src + tar)[14:] |
|
349
|
|
|
concat_comp2 = lzma.compress(tar + src)[14:] |
|
350
|
|
|
else: |
|
351
|
|
|
raise ValueError('Install the PylibLZMA module in order to use lzma ' + |
|
352
|
|
|
'compression similarity') |
|
353
|
|
|
|
|
354
|
|
|
return ((min(len(concat_comp), len(concat_comp2)) - |
|
355
|
|
|
min(len(src_comp), len(tar_comp))) / |
|
356
|
|
|
max(len(src_comp), len(tar_comp))) |
|
357
|
|
|
|
|
358
|
|
|
|
|
359
|
|
|
def sim_ncd_lzma(src, tar): |
|
360
|
|
|
"""Return the NCD similarity between two strings using lzma compression. |
|
361
|
|
|
|
|
362
|
|
|
Normalized compression distance (NCD) :cite:`Cilibrasi:2005`. |
|
363
|
|
|
|
|
364
|
|
|
:param str src: source string for comparison |
|
365
|
|
|
:param str tar: target string for comparison |
|
366
|
|
|
:returns: compression similarity |
|
367
|
|
|
:rtype: float |
|
368
|
|
|
|
|
369
|
|
|
>>> sim_ncd_lzma('cat', 'hat') |
|
370
|
|
|
0.9130434782608696 |
|
371
|
|
|
>>> sim_ncd_lzma('Niall', 'Neil') |
|
372
|
|
|
0.84 |
|
373
|
|
|
>>> sim_ncd_lzma('aluminum', 'Catalan') |
|
374
|
|
|
0.84 |
|
375
|
|
|
>>> sim_ncd_lzma('ATCG', 'TAGC') |
|
376
|
|
|
0.9130434782608696 |
|
377
|
|
|
""" |
|
378
|
|
|
return 1 - dist_ncd_lzma(src, tar) |
|
379
|
|
|
|
|
380
|
|
|
|
|
381
|
|
|
if __name__ == '__main__': |
|
382
|
|
|
import doctest |
|
383
|
|
|
doctest.testmod() |
|
384
|
|
|
|