-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathgenerate-data-english.py
331 lines (284 loc) · 14 KB
/
generate-data-english.py
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
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
# generate-data-english.py - Python module to generate synthetic data based on
# English look-up and error tables.
#
# Original work:
# Peter Christen and Dinusha Vatsalan, January-March 2012
#
# Modified work:
# Tymoteusz Strojny, August 2024
# - Converted from Python 2 to Python 3
#
# =============================================================================
#
# This Source Code Form is subject to the terms of the Mozilla Public
# License, v. 2.0. If a copy of the MPL was not distributed with this
# file, You can obtain one at http://mozilla.org/MPL/2.0/.
#
# =============================================================================
#
# Modifications:
# The original code has been modified from its original form.
# It was initially written for Python 2 and has been converted to Python 3.
#
# The modified code remains subject to the terms of the Mozilla Public
# License, v. 2.0.
# =============================================================================
# Import the necessary other modules of the data generator
#
import geco3.basefunctions as basefunctions # Helper functions
import geco3.attrgenfunct as attrgenfunct # Functions to generate independent attribute values
import geco3.contdepfunct as contdepfunct # Functions to generate dependent continuous attribute
# values
import geco3.generator as generator # Main classes to generate records and the data set
import geco3.corruptor as corruptor # Main classes to corrupt attribute values and records
import random
random.seed(42) # Set seed for random generator, so data generation can be
# repeated
# Set the Unicode encoding for this data generation project. This needs to be
# changed to another encoding for different Unicode character sets.
# Valid encoding strings are listed here:
# http://docs.python.org/library/codecs.html#standard-encodings
#
unicode_encoding_used = 'ascii'
# The name of the record identifier attribute (unique value for each record).
# This name cannot be given as name to any other attribute that is generated.
#
rec_id_attr_name = 'rec-id'
# Set the file name of the data set to be generated (this will be a comma
# separated values, CSV, file).
#
out_file_name = 'example-data-english.csv'
# Set how many original and how many duplicate records are to be generated.
#
num_org_rec = 10000
num_dup_rec = 10000
# Set the maximum number of duplicate records can be generated per original
# record.
#
max_duplicate_per_record = 3
# Set the probability distribution used to create the duplicate records for one
# original record (possible values are: 'uniform', 'poisson', 'zipf').
#
num_duplicates_distribution = 'zipf'
# Set the maximum number of modification that can be applied to a single
# attribute (field).
#
max_modification_per_attr = 1
# Set the number of modification that are to be applied to a record.
#
num_modification_per_record = 5
# Check if the given the unicode encoding selected is valid.
#
basefunctions.check_unicode_encoding_exists(unicode_encoding_used)
# -----------------------------------------------------------------------------
# Define the attributes to be generated (using methods from the generator.py
# module).
#
gname_attr = \
generator.GenerateFreqAttribute(attribute_name = 'given-name',
freq_file_name = 'lookup-files/givenname_f_freq.csv',
has_header_line = False,
unicode_encoding = unicode_encoding_used)
sname_attr = \
generator.GenerateFreqAttribute(attribute_name = 'surname',
freq_file_name = 'lookup-files/surname-freq.csv',
has_header_line = False,
unicode_encoding = unicode_encoding_used)
postcode_attr = \
generator.GenerateFreqAttribute(attribute_name = 'postcode',
freq_file_name = 'lookup-files/postcode_act_freq.csv',
has_header_line = False,
unicode_encoding = unicode_encoding_used)
phone_num_attr = \
generator.GenerateFuncAttribute(attribute_name = 'telephone-number',
function = attrgenfunct.generate_phone_number_australia)
credit_card_attr = \
generator.GenerateFuncAttribute(attribute_name = 'credit-card-number',
function = attrgenfunct.generate_credit_card_number)
age_uniform_attr = \
generator.GenerateFuncAttribute(attribute_name = 'age-uniform',
function = attrgenfunct.generate_uniform_age,
parameters = [0,120])
income_normal_attr = \
generator.GenerateFuncAttribute(attribute_name = 'income-normal',
function = attrgenfunct.generate_normal_value,
parameters = [50000,20000, 0, 1000000, 'float2'])
rating_normal_attr = \
generator.GenerateFuncAttribute(attribute_name = 'rating-normal',
function = attrgenfunct.generate_normal_value,
parameters = [0.0,1.0, None, None, 'float9'])
gender_city_comp_attr = \
generator.GenerateCateCateCompoundAttribute(\
categorical1_attribute_name = 'gender',
categorical2_attribute_name = 'city',
lookup_file_name = 'lookup-files/gender-city.csv',
has_header_line = True,
unicode_encoding = 'ascii')
sex_income_comp_attr = \
generator.GenerateCateContCompoundAttribute(\
categorical_attribute_name = 'sex',
continuous_attribute_name = 'income',
continuous_value_type = 'float1',
lookup_file_name = 'lookup-files/gender-income.csv',
has_header_line = False,
unicode_encoding = 'ascii')
gender_town_salary_comp_attr = \
generator.GenerateCateCateContCompoundAttribute(\
categorical1_attribute_name = 'alternative-gender',
categorical2_attribute_name = 'town',
continuous_attribute_name = 'salary',
continuous_value_type = 'float4',
lookup_file_name = 'lookup-files/gender-city-income.csv',
has_header_line = False,
unicode_encoding = 'ascii')
age_blood_pressure_comp_attr = \
generator.GenerateContContCompoundAttribute(\
continuous1_attribute_name = 'age',
continuous2_attribute_name = 'blood-pressure',
continuous1_funct_name = 'uniform',
continuous1_funct_param = [10,110],
continuous2_function = contdepfunct.blood_pressure_depending_on_age,
continuous1_value_type = 'int',
continuous2_value_type = 'float3')
age_salary_comp_attr = \
generator.GenerateContContCompoundAttribute(\
continuous1_attribute_name = 'age2',
continuous2_attribute_name = 'salary2',
continuous1_funct_name = 'normal',
continuous1_funct_param = [45,20,15,130],
continuous2_function = contdepfunct.salary_depending_on_age,
continuous1_value_type = 'int',
continuous2_value_type = 'float1')
# -----------------------------------------------------------------------------
# Define how the generated records are to be corrupted (using methods from
# the corruptor.py module).
# For a value edit corruptor, the sum or the four probabilities given must
# be 1.0.
#
edit_corruptor = \
corruptor.CorruptValueEdit(\
position_function = corruptor.position_mod_normal,
char_set_funct = basefunctions.char_set_ascii,
insert_prob = 0.5,
delete_prob = 0.5,
substitute_prob = 0.0,
transpose_prob = 0.0)
edit_corruptor2 = \
corruptor.CorruptValueEdit(\
position_function = corruptor.position_mod_uniform,
char_set_funct = basefunctions.char_set_ascii,
insert_prob = 0.25,
delete_prob = 0.25,
substitute_prob = 0.25,
transpose_prob = 0.25)
surname_misspell_corruptor = \
corruptor.CorruptCategoricalValue(\
lookup_file_name = 'lookup-files/surname-misspell.csv',
has_header_line = False,
unicode_encoding = unicode_encoding_used)
ocr_corruptor = corruptor.CorruptValueOCR(\
position_function = corruptor.position_mod_normal,
lookup_file_name = 'lookup-files/ocr-variations.csv',
has_header_line = False,
unicode_encoding = unicode_encoding_used)
keyboard_corruptor = corruptor.CorruptValueKeyboard(\
position_function = corruptor.position_mod_normal,
row_prob = 0.5,
col_prob = 0.5)
phonetic_corruptor = corruptor.CorruptValuePhonetic(\
lookup_file_name = 'lookup-files/phonetic-variations.csv',
has_header_line = False,
unicode_encoding = unicode_encoding_used)
missing_val_corruptor = corruptor.CorruptMissingValue()
postcode_missing_val_corruptor = corruptor.CorruptMissingValue(\
missing_val='missing')
given_name_missing_val_corruptor = corruptor.CorruptMissingValue(\
missing_value='unknown')
# -----------------------------------------------------------------------------
# Define the attributes to be generated for this data set, and the data set
# itself.
#
attr_name_list = ['gender', 'given-name', 'surname', 'postcode', 'city',
'telephone-number', 'credit-card-number', 'income-normal',
'age-uniform', 'income', 'age', 'sex', 'blood-pressure']
attr_data_list = [gname_attr, sname_attr, postcode_attr, phone_num_attr,
credit_card_attr, age_uniform_attr, income_normal_attr,
gender_city_comp_attr, sex_income_comp_attr,
gender_town_salary_comp_attr, age_blood_pressure_comp_attr,
age_salary_comp_attr]
# Nothing to change here - set-up the data set generation object.
#
test_data_generator = generator.GenerateDataSet(output_file_name = \
out_file_name,
write_header_line = True,
rec_id_attr_name = rec_id_attr_name,
number_of_records = num_org_rec,
attribute_name_list = attr_name_list,
attribute_data_list = attr_data_list,
unicode_encoding = \
unicode_encoding_used)
# Define the probability distribution of how likely an attribute will be
# selected for a modification.
# Each of the given probability values must be between 0 and 1, and the sum of
# them must be 1.0.
# If a probability is set to 0 for a certain attribute, then no modification
# will be applied on this attribute.
#
attr_mod_prob_dictionary = {'gender':0.1, 'given-name':0.2,'surname':0.2,
'postcode':0.1,'city':0.1, 'telephone-number':0.15,
'credit-card-number':0.1,'age':0.05}
# Define the actual corruption (modification) methods that will be applied on
# the different attributes.
# For each attribute, the sum of probabilities given must sum to 1.0.
#
attr_mod_data_dictionary = {'gender':[(1.0, missing_val_corruptor)],
'surname':[(0.1, surname_misspell_corruptor),
(0.1, ocr_corruptor),
(0.1, keyboard_corruptor),
(0.7, phonetic_corruptor)],
'given-name':[(0.1, edit_corruptor2),
(0.1, ocr_corruptor),
(0.1, keyboard_corruptor),
(0.7, phonetic_corruptor)],
'postcode':[(0.8, keyboard_corruptor),
(0.2, postcode_missing_val_corruptor)],
'city':[(0.1, edit_corruptor),
(0.1, missing_val_corruptor),
(0.4, keyboard_corruptor),
(0.4, phonetic_corruptor)],
'age':[(1.0, edit_corruptor2)],
'telephone-number':[(1.0, missing_val_corruptor)],
'credit-card-number':[(1.0, edit_corruptor)]}
# Nothing to change here - set-up the data set corruption object
#
test_data_corruptor = corruptor.CorruptDataSet(number_of_org_records = \
num_org_rec,
number_of_mod_records = num_dup_rec,
attribute_name_list = attr_name_list,
max_num_dup_per_rec = \
max_duplicate_per_record,
num_dup_dist = \
num_duplicates_distribution,
max_num_mod_per_attr = \
max_modification_per_attr,
num_mod_per_rec = \
num_modification_per_record,
attr_mod_prob_dict = \
attr_mod_prob_dictionary,
attr_mod_data_dict = \
attr_mod_data_dictionary)
# =============================================================================
# No need to change anything below here
# Start the data generation process
#
rec_dict = test_data_generator.generate()
assert len(rec_dict) == num_org_rec # Check the number of generated records
# Corrupt (modify) the original records into duplicate records
#
rec_dict = test_data_corruptor.corrupt_records(rec_dict)
assert len(rec_dict) == num_org_rec+num_dup_rec # Check total number of records
# Write generate data into a file
#
test_data_generator.write()
# End.
# =============================================================================