-
Notifications
You must be signed in to change notification settings - Fork 0
/
graph.py
738 lines (611 loc) · 30.4 KB
/
graph.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
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
import pandas as pd
import pickle
import re
import nltk
from nltk.corpus import stopwords
import math
from node_embeddings import *
from typing import Union
from math import isinf
from time import time
def isNaN(num):
"""
___DEPRECATED___
"""
return num != num
def get_order_of_magnitude(number: float) -> int:
"""Compute the order of magnitude of a number
Args:
number (float): the input number
Returns:
int: the order of magnitude of the number
"""
if number == 0:
return 0 # Logarithm of 0 is undefined, return 0 as the order of magnitude
else:
return int(math.floor(math.log10(abs(number))))
def is_float(string: str) -> bool:
"""Tells if a string represents a float number
Args:
string (str): the input string
Returns:
bool: True if the string represents a float
"""
try:
float(string)
return True
except ValueError:
return False
def preprocess_numbers(n: Union[float, str], operations: list=['cast_to_float','discretize_strict']) -> str:
"""Preprocessing operations for numbers are performed
Args:
n (Union[float, str]): the number to preprocess
operations (list, optional): the list of preprocessing operation to perform. Defaults to ['cast_to_float', 'discretize_strict'].
Returns:
str: the preprocessed number as a string
"""
if 'cast_to_float' in operations:
n = float(n)
if isinf(n):
return str(n)
if 'discretize_strict' in operations:
div = 10 ** get_order_of_magnitude(n)
n = n//div*div
return str(n)
class String_token_preprocessor:
def __init__(self, language: str='english') -> None:
"""The class init method
Args:
language (str, optional): the language of the strings to preprocess. Defaults to 'english'.
"""
nltk.download('stopwords')
self.stopwords = stopwords.words(language)
def __call__(self, s: str, token_length_limit: int=None, operations: list=['lowercase', 'split', 'remove_stop_words']) -> str:
"""The call method that is used to preprocess strings
Args:
s (str): the string to preprocess
token_length_limit (int, optional): parameter that limit the number of words of the string and cause a truncate operation. Defaults to None.
operations (list, optional): the preprocessing operation to perform. Defaults to ['lowercase', 'split', 'remove_stop_words'].
Returns:
str: the preprocessed string
"""
out = s
if len(operations) == 0:
return [out]
if 'lowercase' in operations:
out = out.lower()
if 'split' in operations:
out = re.split(' |_|\|', out)
if 'remove_stop_words' in operations:
out = [t for t in out if not(t in self.stopwords)]
if token_length_limit:
out = out[0:token_length_limit]
return out
class Graph:
def get_number_of_nodes(self) -> int:
"""Provides the number of nodes in the graph
Returns:
int: the number of nodes
"""
return len(self.index_to_token)
def __str__(self) -> str:
"""str method
Returns:
str: the string representation of the graph
"""
return ''.join(f'{self.index_to_token[self.edges[0][i]]}<-->{self.index_to_token[self.edges[1][i]]}\n' for i in range(self.number_of_edges))
def __add_edge(self, id_a: int, id_b: int) -> None:
"""Add a new edge to the graph provided 2 nodes
Args:
id_a (int): index of the first node
id_b (int): index of the second node
"""
self.edges[0].append(id_a)
self.edges[1].append(id_b)
self.number_of_edges += 1
self.edges[0].append(id_b)
self.edges[1].append(id_a)
self.number_of_edges += 1
def __get_next_index(self, category: str) -> int:
"""Provide an univocal index for the specified category of node
Args:
category (str): the category of the node {'column', 'row', 'value'}
Raises:
Exception: it is raised if the category format is not supported
Returns:
int: the new index
"""
if category=='column':
out = self.next_column_index
self.next_column_index+=1
elif category=='row':
out = self.next_row_index
self.next_row_index+=1
elif category=='value':
out = self.next_value_index
self.next_value_index+=1
else:
raise Exception('Unexpected index format')
return out
def __add_value_to_index(self, value_index: int, column_idx: int, row_idx: int) -> None:
"""Indicize the identifier of a new cell
Args:
value_index (int): the index of the cell
column_idx (int): index of its column
row_idx (int): index of its row
"""
self.columns_rows_to_values[column_idx].append(value_index)
self.columns_rows_to_values[row_idx].append(value_index)
def __generate_feature_matrix(self, embeddings: torch.Tensor) -> torch.Tensor:
"""Generate the feature matrix containing the initial embeddings of the nodes
Args:
embeddings (torch.Tensor): the embeddings of the cell nodes
Returns:
torch.Tensor: the feature matrix
"""
out = [torch.mean(embeddings[l], dim=0).reshape(1,-1) for l in self.columns_rows_to_values]
out = torch.cat(out, dim=0) #cat of a list
out = torch.cat((out, embeddings), dim=0)
return out
def __init__(self, df: pd.DataFrame, table_name: str, embedding_buffer_type: str='sha256', embedding_buffer: Embedding_buffer=None, preprocess_string_token: String_token_preprocessor=None,
token_length_limit: int=1000,link_tuple_token: bool=True, link_token_attribute: bool=True, link_tuple_attribute: bool=False,
attribute_preprocess_operations: list=['lowercase', 'drop_numbers_from_strings'],
string_preprocess_operations: list=['lowercase', 'split', 'remove_stop_words'],
number_preprocess_operations: list=['cast_to_float'], drop_na: bool=False, verbose: bool=False,
merge_nodes_same_value: bool=True) -> None:
if embedding_buffer_type == 'sha256':
self.init_sha256(df=df, table_name=table_name,embedding_buffer=embedding_buffer, link_tuple_token=link_tuple_token, merge_nodes_same_value=merge_nodes_same_value,
link_token_attribute=link_token_attribute, link_tuple_attribute=link_tuple_attribute, drop_na=drop_na, verbose=verbose)
elif embedding_buffer_type == 'fasttext':
self.init_fasttext(df=df, table_name=table_name, embedding_buffer=embedding_buffer, preprocess_string_token=preprocess_string_token,
token_length_limit=token_length_limit, link_tuple_token=link_tuple_token, link_token_attribute=link_token_attribute,
link_tuple_attribute=link_tuple_attribute, attribute_preprocess_operations=attribute_preprocess_operations, string_preprocess_operations=string_preprocess_operations,
number_preprocess_operations=number_preprocess_operations, drop_na=drop_na, verbose=verbose)
def init_sha256(self, df: pd.DataFrame, table_name: str, embedding_buffer: Embedding_buffer=None,
link_tuple_token: bool=True, link_token_attribute: bool=True, link_tuple_attribute: bool=False,
drop_na: bool=False, verbose: bool=False, merge_nodes_same_value: bool=False) -> None:
"""A dataframe will be processed to generate nodes and edges to add to the graph
Args:
df (pd.DataFrame): the dataframe to process
table_name (str): the name of the dataframe, it will be used during the node generation
embedding_buffer (Embedding_buffer, optional): object part of the Embedding_buffer class. Defaults to None.
link_tuple_token (bool, optional): if true tuples and tokens will be linked by edges. Defaults to True.
link_token_attribute (bool, optional): if true tokens and attributes will be linked by edges. Defaults to True.
link_tuple_attribute (bool, optional): if true tuples and attributes will be linked by edges. Defaults to False.
drop_na (bool, optional): set to True to drop all the nan and the nan axises. Defaults to False.
verbose (bool, optional): set True to print debug stuff. Defaults to False.
Raises:
Exception: it occurs if the provided dataframe is empty
Exception: it occurs if a token of an unsupported type appears
"""
self.edges = [[],[]]
self.X = None
self.table_name = table_name
self.number_of_edges = 0
if embedding_buffer == None:
embedding_buffer = Hash_embedding_buffer()
link_tuple_token = link_tuple_token
link_token_attribute = link_token_attribute
link_tuple_attribute = link_tuple_attribute
if drop_na:
df.dropna(axis=0,how='all', inplace=True)
df.dropna(axis=1,how='all', inplace=True)
n_columns = df.shape[1]
n_rows = df.shape[0]
if (n_columns == 0) or (n_rows == 0):
raise Exception('You cannot generate a graph from an empty DataFrame')
self.next_column_index = 0
self.next_row_index = n_columns
self.next_value_index = n_columns + n_rows
self.columns_rows_to_values = [[] for _ in range(n_columns+n_rows)] #contains for every column and row the list of the indexes of the associted values
index_left_shift = len(self.columns_rows_to_values)
column_indexes = [i for i in range(n_columns)]
value_to_index = {}
values_count = 0
#Tuple and token node
for i in range(df.shape[0]):
row_index = self.__get_next_index('row')
if (i % 100 == 0) and verbose:
print(f'Row: {i}/{n_rows}')
if link_tuple_attribute:
for id in column_indexes:
self.__add_edge(row_index, id)
for j in range(df.shape[1]):
t = df.iloc[i,j]
#NaN values management
if pd.isnull(t):
sentence = 'NULL'
sentence = str(t)
if merge_nodes_same_value:
try:
value_index = value_to_index[sentence]
self.__add_value_to_index(value_index-index_left_shift, j, row_index)
except:
embedding_buffer(sentence)
value_index = self.__get_next_index('value')
self.__add_value_to_index(values_count, j, row_index)
values_count += 1
value_to_index[sentence] = value_index
else:
embedding_buffer(sentence)
value_index = self.__get_next_index('value')
self.__add_value_to_index(values_count, j, row_index)
values_count += 1
value_to_index[sentence] = value_index
if link_tuple_token:
self.__add_edge(value_index, row_index)
if link_token_attribute:
self.__add_edge(value_index, column_indexes[j])
value_embeddings = embedding_buffer.pop_embeddings()
self.X = self.__generate_feature_matrix(value_embeddings)
#device = 'cuda' if torch.cuda.is_available() else 'cpu'
self.edges = torch.tensor(self.edges, dtype=torch.long)#.to(device=device)
def init_fasttext(self, df: pd.DataFrame, table_name: str, embedding_buffer: Embedding_buffer=None, preprocess_string_token: String_token_preprocessor=None,
token_length_limit: int=1000,link_tuple_token: bool=True, link_token_attribute: bool=True, link_tuple_attribute: bool=False,
attribute_preprocess_operations: list=['lowercase', 'drop_numbers_from_strings'],
string_preprocess_operations: list=['lowercase', 'split', 'remove_stop_words'],
number_preprocess_operations: list=['cast_to_float'], drop_na: bool=False, verbose: bool=False) -> None:
"""A dataframe will be processed to generate nodes and edges to add to the graph
Args:
df (pd.DataFrame): the dataframe to process
table_name (str): the name of the dataframe, it will be used during the node generation
embedding_buffer (Embedding_buffer): an object of class Embedding_Buffer
preprocess_string_token (String_token_preprocessor): an object of type String_token_preprocessor
token_length_limit (int, optional): the max length acceptable for the sentences. Defaults to 20.
link_tuple_token (bool, optional): if true tuples and tokens will be linked by edges. Defaults to True.
link_token_attribute (bool, optional): if true tokens and attributes will be linked by edges. Defaults to True.
link_tuple_attribute (bool, optional): if true tuples and attributes will be linked by edges. Defaults to False.
attribute_preprocess_operations (list, optional): list of preprocessing operations for attributes. Defaults to ['lowercase', 'drop_numbers_from_strings'].
string_preprocess_operations (list, optional): list of preprocessing operations for strings. Defaults to ['lowercase', 'split', 'remove_stop_words'].
number_preprocess_operations (list, optional): list of preprocessing operations for numbers. Defaults to ['cast_to_float', 'discretize_strict'].
drop_na (bool, optional): set to True to drop all the nan and the nan axises. Defaults to False.
verbose (bool, optional): set True to print debug stuff. Defaults to False.
Raises:
Exception: it occurs if the provided dataframe is empty
Exception: it occurs if a token of an unsupported type appears
"""
token_length_limit = 1000
self.edges = [[],[]]
self.X = None
self.table_name = table_name
self.number_of_edges = 0
link_tuple_token = link_tuple_token
link_token_attribute = link_token_attribute
link_tuple_attribute = link_tuple_attribute
string_preprocess_operations = ['lowercase', 'split', 'remove_stop_words']
number_preprocess_operations = ['cast_to_float', 'discretize_strict']
if drop_na:
df.dropna(axis=0,how='all', inplace=True)
df.dropna(axis=1,how='all', inplace=True)
n_columns = df.shape[1]
n_rows = df.shape[0]
if (n_columns == 0) or (n_rows == 0):
raise Exception('You cannot generate a graph from an empty DataFrame')
self.next_column_index = 0
self.next_row_index = n_columns
self.next_value_index = n_columns + n_rows
self.columns_rows_to_values = [[] for _ in range(n_columns+n_rows)] #contains for every column and row the list of the indexes of the associted values
index_left_shift = len(self.columns_rows_to_values)
column_indexes = [i for i in range(n_columns)]
value_to_index = {}
values_count = 0
#Tuple and token node
for i in range(df.shape[0]):
row_index = self.__get_next_index('row')
if (i % 100 == 0) and verbose:
print(f'Row: {i}/{n_rows}')
if link_tuple_attribute:
for id in column_indexes:
self.__add_edge(row_index, id)
for j in range(df.shape[1]):
t = df.iloc[i,j]
#NaN values management
if pd.isnull(t):
sentence = '#£$/' #Random key for NaN
try:
value_index = value_to_index[sentence]
self.__add_value_to_index(value_index-index_left_shift, j, row_index)
except:
embedding_buffer.add_nan_embedding()
value_index = self.__get_next_index('value')
self.__add_value_to_index(values_count, j, row_index)
values_count += 1
value_to_index[sentence] = value_index
if link_tuple_token:
self.__add_edge(value_index, row_index)
if link_token_attribute:
self.__add_edge(value_index, column_indexes[j])
continue
if isinstance(t, str) and not(is_float(t)):
token_list = preprocess_string_token(t, token_length_limit,operations=string_preprocess_operations)
elif is_float(str(t)):
#Note: the strings "infinity","Infinity","Inf", and "inf" will trigger an exception and will be skipped
try:
token_list = [preprocess_numbers(t, operations=number_preprocess_operations)]
except:
try:
sentence = '#???$/'
try:
value_index = value_to_index[sentence]
self.__add_value_to_index(value_index-index_left_shift, j, row_index)
except:
embedding_buffer.add_random_embedding()
value_index = self.__get_next_index('value')
self.__add_value_to_index(values_count, j, row_index)
values_count += 1
value_to_index[sentence] = value_index
if link_tuple_token:
self.__add_edge(value_index, row_index)
if link_token_attribute:
self.__add_edge(value_index, column_indexes[j])
continue
except:
print('An exception occurred')
continue
else:
raise Exception(f'The token {t} is of type {type(t)} and it is not supported')
sentence = ' '.join(token_list)
try:
value_index = value_to_index[sentence]
self.__add_value_to_index(value_index-index_left_shift, j, row_index)
except:
embedding_buffer(sentence)
value_index = self.__get_next_index('value')
self.__add_value_to_index(values_count, j, row_index)
values_count += 1
value_to_index[sentence] = value_index
if link_tuple_token:
self.__add_edge(value_index, row_index)
if link_token_attribute:
self.__add_edge(value_index, column_indexes[j])
value_embeddings = embedding_buffer.pop_embeddings()
self.X = self.__generate_feature_matrix(value_embeddings)
#device = 'cuda' if torch.cuda.is_available() else 'cpu'
self.edges = torch.tensor(self.edges, dtype=torch.long)#.to(device=device)
class Graph_Hashed_Node_Embs:
def get_number_of_nodes(self) -> int:
"""Provides the number of nodes in the graph
Returns:
int: the number of nodes
"""
return len(self.index_to_token)
def __str__(self) -> str:
"""str method
Returns:
str: the string representation of the graph
"""
return ''.join(f'{self.index_to_token[self.edges[0][i]]}<-->{self.index_to_token[self.edges[1][i]]}\n' for i in range(self.number_of_edges))
def __add_edge(self, id_a: int, id_b: int) -> None:
"""Add a new edge to the graph provided 2 nodes
Args:
id_a (int): index of the first node
id_b (int): index of the second node
"""
self.edges[0].append(id_a)
self.edges[1].append(id_b)
self.number_of_edges += 1
self.edges[0].append(id_b)
self.edges[1].append(id_a)
self.number_of_edges += 1
def __get_next_index(self, category: str) -> int:
"""Provide an univocal index for the specified category of node
Args:
category (str): the category of the node {'column', 'row', 'value'}
Raises:
Exception: it is raised if the category format is not supported
Returns:
int: the new index
"""
if category=='column':
out = self.next_column_index
self.next_column_index+=1
elif category=='row':
out = self.next_row_index
self.next_row_index+=1
elif category=='value':
out = self.next_value_index
self.next_value_index+=1
else:
raise Exception('Unexpected index format')
return out
def __add_value_to_index(self, value_index: int, column_idx: int, row_idx: int) -> None:
"""Indicize the identifier of a new cell
Args:
value_index (int): the index of the cell
column_idx (int): index of its column
row_idx (int): index of its row
"""
self.columns_rows_to_values[column_idx].append(value_index)
self.columns_rows_to_values[row_idx].append(value_index)
def __generate_feature_matrix(self, embeddings: torch.Tensor) -> torch.Tensor:
"""Generate the feature matrix containing the initial embeddings of the nodes
Args:
embeddings (torch.Tensor): the embeddings of the cell nodes
Returns:
torch.Tensor: the feature matrix
"""
out = [torch.mean(embeddings[l], dim=0).reshape(1,-1) for l in self.columns_rows_to_values]
out = torch.cat(out, dim=0) #cat of a list
out = torch.cat((out, embeddings), dim=0)
return out
def __init__(self, df: pd.DataFrame, table_name: str, embedding_buffer: Embedding_buffer=None,
link_tuple_token: bool=True, link_token_attribute: bool=True, link_tuple_attribute: bool=False,
drop_na: bool=False, verbose: bool=False) -> None:
"""A dataframe will be processed to generate nodes and edges to add to the graph
Args:
df (pd.DataFrame): the dataframe to process
table_name (str): the name of the dataframe, it will be used during the node generation
embedding_buffer (Embedding_buffer, optional): object part of the Embedding_buffer class. Defaults to None.
link_tuple_token (bool, optional): if true tuples and tokens will be linked by edges. Defaults to True.
link_token_attribute (bool, optional): if true tokens and attributes will be linked by edges. Defaults to True.
link_tuple_attribute (bool, optional): if true tuples and attributes will be linked by edges. Defaults to False.
drop_na (bool, optional): set to True to drop all the nan and the nan axises. Defaults to False.
verbose (bool, optional): set True to print debug stuff. Defaults to False.
Raises:
Exception: it occurs if the provided dataframe is empty
Exception: it occurs if a token of an unsupported type appears
"""
self.edges = [[],[]]
self.X = None
self.table_name = table_name
self.number_of_edges = 0
if embedding_buffer == None:
embedding_buffer = Hash_embedding_buffer()
link_tuple_token = link_tuple_token
link_token_attribute = link_token_attribute
link_tuple_attribute = link_tuple_attribute
if drop_na:
df.dropna(axis=0,how='all', inplace=True)
df.dropna(axis=1,how='all', inplace=True)
n_columns = df.shape[1]
n_rows = df.shape[0]
if (n_columns == 0) or (n_rows == 0):
raise Exception('You cannot generate a graph from an empty DataFrame')
self.next_column_index = 0
self.next_row_index = n_columns
self.next_value_index = n_columns + n_rows
self.columns_rows_to_values = [[] for _ in range(n_columns+n_rows)] #contains for every column and row the list of the indexes of the associted values
index_left_shift = len(self.columns_rows_to_values)
column_indexes = [i for i in range(n_columns)]
value_to_index = {}
values_count = 0
#Tuple and token node
for i in range(df.shape[0]):
row_index = self.__get_next_index('row')
if (i % 100 == 0) and verbose:
print(f'Row: {i}/{n_rows}')
if link_tuple_attribute:
for id in column_indexes:
self.__add_edge(row_index, id)
for j in range(df.shape[1]):
t = df.iloc[i,j]
#NaN values management
if pd.isnull(t):
sentence = '#£$/' #Random key for NaN
try:
value_index = value_to_index[sentence]
self.__add_value_to_index(value_index-index_left_shift, j, row_index)
except:
embedding_buffer.add_nan_embedding()
value_index = self.__get_next_index('value')
self.__add_value_to_index(values_count, j, row_index)
values_count += 1
value_to_index[sentence] = value_index
if link_tuple_token:
self.__add_edge(value_index, row_index)
if link_token_attribute:
self.__add_edge(value_index, column_indexes[j])
continue
sentence = str(t)
#Note: the strings "infinity","Infinity","Inf", and "inf" will trigger an exception and will be skipped
try:
value_index = value_to_index[sentence]
self.__add_value_to_index(value_index-index_left_shift, j, row_index)
except:
embedding_buffer(sentence)
value_index = self.__get_next_index('value')
self.__add_value_to_index(values_count, j, row_index)
values_count += 1
value_to_index[sentence] = value_index
if link_tuple_token:
self.__add_edge(value_index, row_index)
if link_token_attribute:
self.__add_edge(value_index, column_indexes[j])
value_embeddings = embedding_buffer.pop_embeddings()
self.X = self.__generate_feature_matrix(value_embeddings)
#device = 'cuda' if torch.cuda.is_available() else 'cpu'
self.edges = torch.tensor(self.edges, dtype=torch.long)#.to(device=device)
class Graph_list(torch.utils.data.Dataset):
def __init__(self, directory_name: str=False,save: bool=False, load: bool=False) -> None:
"""The class init method
Args:
directory_name (str, optional): directory that contains the list. Defaults to False.
save (bool, optional): NA. Defaults to False.
load (bool, optional): NA. Defaults to False.
"""
if directory_name:
self.load(directory_name)
else:
self.graph_list = []
self.size = 0
def __len__(self) -> int:
"""classic len method
Returns:
int: length of the list
"""
return self.size
def __getitem__(self, idx: int) -> Graph:
"""Getitem method
Args:
idx (int): index to retrieve
Raises:
Exception: raised if the index is out of bound
Returns:
Graph: the graph corresponding to the provided index
"""
try:
return self.graph_list[idx]
except IndexError:
raise Exception("Index out of bound")
def add_item(self, g: Graph) -> None:
"""Add a graph to the collection
Args:
g (Graph): the graph to add to the collection
"""
self.graph_list.append(g)
self.size += 1
def load(self, directory_name: str) -> None:
"""Replace the current content with some read from a file
Args:
directory_name (str): path to the graphlist
Raises:
Exception: raised if there is a failure during the read operation
"""
try:
f1 = open(directory_name+'/graph_list.pkl', 'rb')
self.graph_list = pickle.load(f1)
self.size = len(self.graph_list)
f1.close()
except:
raise Exception('Read operation failed')
self.size = len(self.graph_list)
def save(self, directory_name: str) -> None:
"""To save the list
Args:
directory_name (str): directory where to save the necessary files
Raises:
Exception: raised if there is a failure during the write operation
"""
try:
f1 = open(directory_name+'/graph_list.pkl', 'wb')
pickle.dump(self.graph_list, f1)
f1.close()
except:
raise Exception('Write operation failed')
if __name__ == "__main__":
# from syntheticDatasetGenerator import load_test_training_stuff
# data = load_test_training_stuff("/home/francesco.pugnaloni/tmp/small_tables")
# t = data['tables']['25']
# t['1'] = [pd.NA, 15, pd.NA]
# #embedding_buffer = FasttextEmbeddingBuffer(model='fasttext-wiki-news-subwords-300')
# #embedding_buffer = FasttextEmbeddingBuffer()
# embedding_buffer = Bert_Embedding_Buffer()
# string_token_preprocessor = String_token_preprocessor()
# g = Graph(t, 'luca', embedding_buffer, string_token_preprocessor)
# print(f'Number of NA: {torch.sum(torch.isnan(g.X))}')
# print('ok')
mm = {
'a':[1,2,3,4,5,6,7,8],
'b':[9,8,7,6,5,4,3,2],
'c':[9,8,7,6,5,4,3,2]
}
dd = pd.DataFrame(mm)
start = time()
gg = Graph(dd, 'ff', 'sha256')
end = time()
print(f'T_exec: {end-start} sec')
dd = pd.DataFrame(mm)
start = time()
grg = Graph(dd, 'ff', 'sha256', merge_nodes_same_value=False)
end = time()
print(f'T_exec: {end-start} sec')