-
Notifications
You must be signed in to change notification settings - Fork 3
Expand file tree
/
Copy pathLEMMA_experiment.py
More file actions
executable file
·184 lines (155 loc) · 5.47 KB
/
LEMMA_experiment.py
File metadata and controls
executable file
·184 lines (155 loc) · 5.47 KB
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
from ConstrainAgent.ConstrainAgent import ConstrainNormalAgent
from Utils.CausalDiscovery import (
causal_discovery,
)
from Utils.data import load_Lemma_data
from Utils.visualize import visualize_graph
from Utils.RCA import random_walk_with_restart
from Web_tools import (
split_summary_into_sub_questions,
collect_web_content,
generate_dataset_summary,
)
causal_discovery_algorithm = "pc"
# causal_discovery_algorithm = "Exact-Search"
# causal_discovery_algorithm = "DirectLiNGAM"
system_name = "Product_Review"
theme = "MicroService system about Product Review"
day = 20210517
# system_name = "Cloud_Computing"
# theme = "MicroService system about Cloud Computing"
# day = 20231207
print(f"Loading dataset: {system_name}_{day}...")
data_table, log_dir = load_Lemma_data(system_name, day)
data = data_table.values
labels = data_table.columns.tolist()
print(f"Running {causal_discovery_algorithm} algorithm...")
adjacency_matrix = causal_discovery(data, labels, method=causal_discovery_algorithm)
visualize_graph(adjacency_matrix, labels, f"./image/{system_name}_{day}/PC_graph.png")
count = random_walk_with_restart(adjacency_matrix, len(labels) - 1)
count_label_pairs = list(zip(count, labels))
sorted_pairs = sorted(count_label_pairs, key=lambda x: x[0], reverse=True)
print("\nRanked metrics by importance:")
for count, label in sorted_pairs:
print(f"{label}: {count:.4f}")
print("================================================\n")
print("Running ConstrainAgent...")
constrain_agent = ConstrainNormalAgent(
labels,
theme,
graph_matrix=adjacency_matrix,
causal_discovery_algorithm=causal_discovery_algorithm,
use_reasoning=False,
)
constraint_matrix = constrain_agent.run(
use_cache=False,
cache_path=f"./cache/Domain_knowledge/{system_name}_{day}/{causal_discovery_algorithm}",
)
adjacency_matrix_optimized = causal_discovery(
data,
labels,
method=causal_discovery_algorithm,
constraint_matrix=constraint_matrix,
)
visualize_graph(
adjacency_matrix_optimized,
labels,
f"./image/{system_name}_{day}/PC_graph_Optimized.png",
)
count = random_walk_with_restart(adjacency_matrix, len(labels) - 1)
count_label_pairs = list(zip(count, labels))
sorted_pairs = sorted(count_label_pairs, key=lambda x: x[0], reverse=True)
print("\nRanked metrics by importance:")
for count, label in sorted_pairs:
print(f"{label}: {count:.4f}")
print("================================================\n")
print("Running ConstrainAgent with web information...")
# data_info, node_info = split_summary_into_sub_questions(
# open(f"./cache/Summarized_info/{system_name}_{day}_info.txt").read()
# )
collect_web_content(system_name, labels, loop_num=5, use_tavily=False)
data_info, node_info = split_summary_into_sub_questions(
generate_dataset_summary(
system_name,
labels,
output_dir="./cache/Summarized_info",
embeddings_path="./cache/RAG_Database/Embeddings",
save_embeddings=True,
)
)
print(data_info)
print(node_info)
constrain_agent = ConstrainNormalAgent(
labels,
theme,
graph_matrix=adjacency_matrix,
causal_discovery_algorithm=causal_discovery_algorithm,
dataset_information=data_info,
node_information=node_info,
use_reasoning=False,
)
prompt, system_prompt = constrain_agent.domain_knowledge_LLM.generate_prompt(
0, 1, node_info
)
constraint_matrix = constrain_agent.run(
use_cache=False,
cache_path=f"./cache/Domain_knowledge/{system_name}_{day}/{causal_discovery_algorithm}",
)
adjacency_matrix_optimized = causal_discovery(
data,
labels,
method=causal_discovery_algorithm,
constraint_matrix=constraint_matrix,
)
visualize_graph(
adjacency_matrix_optimized,
labels,
f"./image/{system_name}_{day}/PC_graph_Optimized_web.png",
)
count = random_walk_with_restart(adjacency_matrix_optimized, len(labels) - 1)
count_label_pairs = list(zip(count, labels))
sorted_pairs = sorted(count_label_pairs, key=lambda x: x[0], reverse=True)
print("\nRanked metrics by importance:")
for count, label in sorted_pairs:
print(f"{label}: {count:.4f}")
print("================================================\n")
print("Running ConstrainAgent with web information with Reasoning...")
data_info, node_info = split_summary_into_sub_questions(
open(f"./cache/Summarized_info/{system_name}_{day}_info.txt").read()
)
print(data_info)
print(node_info)
constrain_agent = ConstrainNormalAgent(
labels,
theme,
graph_matrix=adjacency_matrix,
causal_discovery_algorithm=causal_discovery_algorithm,
dataset_information=data_info,
node_information=node_info,
use_reasoning=True,
)
prompt, system_prompt = constrain_agent.domain_knowledge_LLM.generate_prompt(
0, 1, node_info
)
constraint_matrix = constrain_agent.run(
use_cache=False,
cache_path=f"./cache/Domain_knowledge/{system_name}_{day}/{causal_discovery_algorithm}",
)
adjacency_matrix_optimized = causal_discovery(
data,
labels,
method=causal_discovery_algorithm,
constraint_matrix=constraint_matrix,
)
visualize_graph(
adjacency_matrix_optimized,
labels,
f"./image/{system_name}_{day}/PC_graph_Optimized_web.png",
)
count = random_walk_with_restart(adjacency_matrix_optimized, len(labels) - 1)
count_label_pairs = list(zip(count, labels))
sorted_pairs = sorted(count_label_pairs, key=lambda x: x[0], reverse=True)
print("\nRanked metrics by importance:")
for count, label in sorted_pairs:
print(f"{label}: {count:.4f}")
print("================================================\n")