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 | class CronQuestions:
def __init__(self):
self.engine = create_engine(DB_CONNECTION_STR)
self.cron_question_dir = DATA_DIR / "CronQuestions" / "questions"
self.cron_kg_dir = DATA_DIR / "CronQuestions" / "kg"
self.id_2_relation = None
self.id_2_entity = None
self.id_alias = None
self.full_df = None
def load_questions(self):
# read pickle file
question_files = [
"test.pickle",
"train.pickle",
"valid.pickle",
]
questions_df = pd.DataFrame()
for question_file in question_files:
with open(self.cron_question_dir / question_file, "rb") as f:
questions = pickle.load(f)
# Convert questions to a DataFrame before appending if not already in DataFrame format
if not isinstance(questions, pd.DataFrame):
questions = pd.DataFrame(questions)
questions_df = pd.concat([questions_df, questions], ignore_index=True)
questions_df.to_csv(self.cron_question_dir / "questions.csv", index=False)
# question,answers,answer_type,template ONLY USE this 4 columns to SQL
questions_df = questions_df[["question", "answer_type", "template", "type"]]
questions_df.to_sql("cron_questions", DB_CONNECTION_STR, if_exists="replace")
logger.info(len(questions_df))
def load_kg(self):
"""
Load the questions to the unified KG table.
:return:
"""
# go to database to check whether we have a table cron_kg exists and with record, if not
# load it from the file
# first check whether the table exists
load_full_kg = False
with self.engine.connect() as conn:
result = conn.execute(
text(
"SELECT COUNT(*) FROM information_schema.tables WHERE table_name = 'cron_kg'"
)
)
cron_kg_table_count = (
result.scalar()
) # Using scalar() to get the first column of the first row
logger.info(cron_kg_table_count)
if cron_kg_table_count != 0:
result = conn.execute(text("SELECT COUNT(*) FROM cron_kg"))
cron_kg_record_count = (
result.scalar()
) # Using scalar() to get the first column of the first row
if cron_kg_record_count == 0:
load_full_kg = True
else:
load_full_kg = True
if load_full_kg is False:
with timer(logger, "read full.csv"):
self.full_df = pd.read_sql("SELECT * FROM cron_kg", self.engine)
return
with timer(logger, "load basic dataset information"):
id_2_entity = pd.read_table(
self.cron_kg_dir / "wd_id2entity_text.txt", header=None
)
id_2_relation = pd.read_table(
self.cron_kg_dir / "wd_id2relation_text.txt", header=None
)
# rename columns to ['id', 'entity']
id_2_entity = id_2_entity.rename(columns={0: "id", 1: "entity"})
# rename columns to ['id', 'relation']
id_2_relation = id_2_relation.rename(columns={0: "id", 1: "relation"})
# Ensure that 'id' is the index
id_2_entity.set_index("id", inplace=True)
id_2_relation.set_index("id", inplace=True)
self.id_2_relation = id_2_relation
self.id_2_entity = id_2_entity
self.id_alias = pd.read_pickle(self.cron_kg_dir / "wd_id_to_aliases.pickle")
with timer(logger, "load full.csv"):
full_df = pd.read_table(self.cron_kg_dir / "full.txt", header=None)
full_df = full_df.rename(
columns={
0: "head",
1: "relation",
2: "tail",
3: "start_year",
4: "end_year",
}
)
# Create mapping series
entity_mapping = id_2_entity["entity"]
relation_mapping = id_2_relation["relation"]
# Convert to nlp triple format
full_df["head"] = full_df["head"].map(entity_mapping)
full_df["relation"] = full_df["relation"].map(relation_mapping)
full_df["tail"] = full_df["tail"].map(entity_mapping)
self.full_df = full_df
# add -01-01 to start_year and end_year
self.full_df["start_year"] = (
self.full_df["start_year"].astype(str) + "-01-01"
)
self.full_df["end_year"] = self.full_df["end_year"].astype(str) + "-01-01"
self.full_df.to_sql("cron_kg", DB_CONNECTION_STR, if_exists="replace")
def unified_kg(self):
"""
Get the cron_kg into the unified kg format
Which include the following columns:
subject
predicate
object
subject_json (json) empty here
object_json (json) empty here
start_time (str)
end_time (str)
"""
cursor = self.engine.connect()
# run sql directly from the cron_kg
cursor.execute(
text(
"""
DO
$$
BEGIN
IF NOT EXISTS (SELECT 1 FROM information_schema.tables WHERE table_name = 'unified_kg_cron') THEN
CREATE TABLE unified_kg_cron(
id SERIAL PRIMARY KEY,
subject TEXT,
subject_json JSON DEFAULT '{}'::JSON,
predicate TEXT,
predicate_json JSON DEFAULT '{}'::JSON,
object TEXT,
object_json JSON DEFAULT '{}'::JSON,
start_time TEXT,
end_time TEXT
);
END IF;
TRUNCATE TABLE unified_kg_cron;
INSERT INTO unified_kg_cron(subject, predicate, object, start_time, end_time)
SELECT head, relation, tail, start_year, end_year FROM cron_kg;
END
$$
"""
)
)
cursor.commit()
cursor.close()
|