\n 📖 This is an example of how to program a one-round conversation in AgentScope.
Modules Used:
- DashScope Chat:
- Each application must contain a model configured. - UserAgent:
- Represents a user in a application. - DialogAgent:
- Agent for dialog in an application.
\n ",
90 | "typenode": false,
91 | "inputs": {},
92 | "outputs": {},
93 | "pos_x": 68,
94 | "pos_y": 67
95 | }
96 | }
97 | }
98 | }
99 | }
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/utils/__init__.py:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/open-sciencelab/Virtual-Scientists/9a6d0f32f9adbccef68f70f1705fa12fc9d10e23/agentscope-main/src/agentscope/utils/__init__.py
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/__init__.py:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/open-sciencelab/Virtual-Scientists/9a6d0f32f9adbccef68f70f1705fa12fc9d10e23/agentscope-main/src/agentscope/web/__init__.py
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/gradio/__init__.py:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/open-sciencelab/Virtual-Scientists/9a6d0f32f9adbccef68f70f1705fa12fc9d10e23/agentscope-main/src/agentscope/web/gradio/__init__.py
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/gradio/constants.py:
--------------------------------------------------------------------------------
1 | # -*- coding: utf-8 -*-
2 | """ Some constants used in the AS studio"""
3 |
4 | _SPEAK = "**speak**"
5 |
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/workstation/__init__.py:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/open-sciencelab/Virtual-Scientists/9a6d0f32f9adbccef68f70f1705fa12fc9d10e23/agentscope-main/src/agentscope/web/workstation/__init__.py
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/workstation/examples/0_two_agent.json:
--------------------------------------------------------------------------------
1 | {
2 | "4": {
3 | "name": "dashscope_chat",
4 | "data": {
5 | "args": {
6 | "config_name": "qwen",
7 | "model_name": "qwen-max",
8 | "api_key": "************",
9 | "temperature": 0.1,
10 | "seed": 1,
11 | "model_type": "dashscope_chat",
12 | "messages_key": "input"
13 | }
14 | },
15 | "inputs": {},
16 | "outputs": {}
17 | },
18 | "5": {
19 | "name": "Message",
20 | "data": {
21 | "args": {
22 | "name": "Host",
23 | "content": "Hello",
24 | "url": ""
25 | }
26 | },
27 | "inputs": {
28 | "input_1": {
29 | "connections": []
30 | }
31 | },
32 | "outputs": {
33 | "output_1": {
34 | "connections": [
35 | {
36 | "node": "7",
37 | "output": "input_1"
38 | }
39 | ]
40 | }
41 | }
42 | },
43 | "6": {
44 | "name": "UserAgent",
45 | "data": {
46 | "args": {
47 | "name": "User"
48 | }
49 | },
50 | "inputs": {
51 | "input_1": {
52 | "connections": [
53 | {
54 | "node": "7",
55 | "input": "output_1"
56 | }
57 | ]
58 | }
59 | },
60 | "outputs": {
61 | "output_1": {
62 | "connections": []
63 | }
64 | }
65 | },
66 | "7": {
67 | "name": "DialogAgent",
68 | "data": {
69 | "args": {
70 | "name": "Alice",
71 | "sys_prompt": "You are Alice.",
72 | "model_config_name": "qwen"
73 | }
74 | },
75 | "inputs": {
76 | "input_1": {
77 | "connections": [
78 | {
79 | "node": "5",
80 | "input": "output_1"
81 | }
82 | ]
83 | }
84 | },
85 | "outputs": {
86 | "output_1": {
87 | "connections": [
88 | {
89 | "node": "6",
90 | "output": "input_1"
91 | }
92 | ]
93 | }
94 | }
95 | }
96 | }
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/workstation/examples/1_conversation.json:
--------------------------------------------------------------------------------
1 | {
2 | "2": {
3 | "name": "dashscope_chat",
4 | "data": {
5 | "args": {
6 | "config_name": "qwen",
7 | "model_name": "qwen-max",
8 | "api_key": "************",
9 | "temperature": 0.1,
10 | "seed": 1,
11 | "model_type": "dashscope_chat",
12 | "messages_key": "input"
13 | }
14 | },
15 | "inputs": {},
16 | "outputs": {}
17 | },
18 | "3": {
19 | "name": "ForLoopPipeline",
20 | "data": {
21 | "elements": [
22 | "4"
23 | ],
24 | "args": {
25 | "max_loop": 3,
26 | "break_func": ""
27 | }
28 | },
29 | "inputs": {
30 | "input_1": {
31 | "connections": [
32 | {
33 | "node": "7",
34 | "input": "output_1"
35 | }
36 | ]
37 | }
38 | },
39 | "outputs": {
40 | "output_1": {
41 | "connections": []
42 | }
43 | }
44 | },
45 | "4": {
46 | "name": "SequentialPipeline",
47 | "data": {
48 | "elements": [
49 | "5",
50 | "6"
51 | ]
52 | },
53 | "inputs": {
54 | "input_1": {
55 | "connections": []
56 | }
57 | },
58 | "outputs": {
59 | "output_1": {
60 | "connections": []
61 | }
62 | }
63 | },
64 | "5": {
65 | "name": "DialogAgent",
66 | "data": {
67 | "args": {
68 | "name": "Alice",
69 | "sys_prompt": "You are Alice.",
70 | "model_config_name": "qwen"
71 | }
72 | },
73 | "inputs": {
74 | "input_1": {
75 | "connections": []
76 | }
77 | },
78 | "outputs": {
79 | "output_1": {
80 | "connections": []
81 | }
82 | }
83 | },
84 | "6": {
85 | "name": "UserAgent",
86 | "data": {
87 | "args": {
88 | "name": "User"
89 | }
90 | },
91 | "inputs": {
92 | "input_1": {
93 | "connections": []
94 | }
95 | },
96 | "outputs": {
97 | "output_1": {
98 | "connections": []
99 | }
100 | }
101 | },
102 | "7": {
103 | "name": "Message",
104 | "data": {
105 | "args": {
106 | "name": "Host",
107 | "content": "Hello",
108 | "url": ""
109 | }
110 | },
111 | "inputs": {
112 | "input_1": {
113 | "connections": []
114 | }
115 | },
116 | "outputs": {
117 | "output_1": {
118 | "connections": [
119 | {
120 | "node": "3",
121 | "output": "input_1"
122 | }
123 | ]
124 | }
125 | }
126 | }
127 | }
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/workstation/examples/2_msghub.json:
--------------------------------------------------------------------------------
1 | {
2 | "2": {
3 | "name": "dashscope_chat",
4 | "data": {
5 | "args": {
6 | "config_name": "qwen",
7 | "model_name": "qwen-max",
8 | "api_key": "************",
9 | "temperature": 0.1,
10 | "seed": 1,
11 | "model_type": "dashscope_chat",
12 | "messages_key": "input"
13 | }
14 | },
15 | "inputs": {},
16 | "outputs": {}
17 | },
18 | "3": {
19 | "name": "MsgHub",
20 | "data": {
21 | "elements": [
22 | "4"
23 | ],
24 | "args": {
25 | "announcement": {
26 | "name": "Host",
27 | "content": "Welcome to group chat!"
28 | }
29 | }
30 | },
31 | "inputs": {
32 | "input_1": {
33 | "connections": [
34 | {
35 | "node": "7",
36 | "input": "output_1"
37 | }
38 | ]
39 | }
40 | },
41 | "outputs": {
42 | "output_1": {
43 | "connections": []
44 | }
45 | }
46 | },
47 | "4": {
48 | "name": "SequentialPipeline",
49 | "data": {
50 | "elements": [
51 | "5",
52 | "6"
53 | ]
54 | },
55 | "inputs": {
56 | "input_1": {
57 | "connections": []
58 | }
59 | },
60 | "outputs": {
61 | "output_1": {
62 | "connections": []
63 | }
64 | }
65 | },
66 | "5": {
67 | "name": "DialogAgent",
68 | "data": {
69 | "args": {
70 | "name": "Alice",
71 | "sys_prompt": "You are Alice.",
72 | "model_config_name": "qwen"
73 | }
74 | },
75 | "inputs": {
76 | "input_1": {
77 | "connections": []
78 | }
79 | },
80 | "outputs": {
81 | "output_1": {
82 | "connections": []
83 | }
84 | }
85 | },
86 | "6": {
87 | "name": "UserAgent",
88 | "data": {
89 | "args": {
90 | "name": "User"
91 | }
92 | },
93 | "inputs": {
94 | "input_1": {
95 | "connections": []
96 | }
97 | },
98 | "outputs": {
99 | "output_1": {
100 | "connections": []
101 | }
102 | }
103 | },
104 | "7": {
105 | "name": "Message",
106 | "data": {
107 | "args": {
108 | "name": "Host",
109 | "content": "Hello",
110 | "url": ""
111 | }
112 | },
113 | "inputs": {
114 | "input_1": {
115 | "connections": []
116 | }
117 | },
118 | "outputs": {
119 | "output_1": {
120 | "connections": [
121 | {
122 | "node": "3",
123 | "output": "input_1"
124 | }
125 | ]
126 | }
127 | }
128 | }
129 | }
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/workstation/examples/4_copiednode.json:
--------------------------------------------------------------------------------
1 | {
2 | "7": {
3 | "name": "DialogAgent",
4 | "data": {
5 | "args": {
6 | "name": "Alice",
7 | "sys_prompt": "You are Alice.",
8 | "model_config_name": "qwen"
9 | }
10 | },
11 | "inputs": {
12 | "input_1": {
13 | "connections": [
14 | {
15 | "node": "11",
16 | "input": "output_1"
17 | }
18 | ]
19 | }
20 | },
21 | "outputs": {
22 | "output_1": {
23 | "connections": [
24 | {
25 | "node": "8",
26 | "output": "input_1"
27 | }
28 | ]
29 | }
30 | }
31 | },
32 | "8": {
33 | "name": "UserAgent",
34 | "data": {
35 | "args": {
36 | "name": "User"
37 | }
38 | },
39 | "inputs": {
40 | "input_1": {
41 | "connections": [
42 | {
43 | "node": "7",
44 | "input": "output_1"
45 | }
46 | ]
47 | }
48 | },
49 | "outputs": {
50 | "output_1": {
51 | "connections": [
52 | {
53 | "node": "12",
54 | "output": "input_1"
55 | }
56 | ]
57 | }
58 | }
59 | },
60 | "10": {
61 | "name": "dashscope_chat",
62 | "data": {
63 | "args": {
64 | "config_name": "qwen",
65 | "model_name": "qwen-max",
66 | "api_key": "************",
67 | "temperature": 0.1,
68 | "seed": 1,
69 | "model_type": "dashscope_chat",
70 | "messages_key": "input"
71 | }
72 | },
73 | "inputs": {},
74 | "outputs": {}
75 | },
76 | "11": {
77 | "name": "Message",
78 | "data": {
79 | "args": {
80 | "name": "Host",
81 | "content": "Hi"
82 | }
83 | },
84 | "inputs": {
85 | "input_1": {
86 | "connections": []
87 | }
88 | },
89 | "outputs": {
90 | "output_1": {
91 | "connections": [
92 | {
93 | "node": "7",
94 | "output": "input_1"
95 | }
96 | ]
97 | }
98 | }
99 | },
100 | "12": {
101 | "name": "CopyNode",
102 | "data": {
103 | "elements": [
104 | "7"
105 | ]
106 | },
107 | "inputs": {
108 | "input_1": {
109 | "connections": [
110 | {
111 | "node": "8",
112 | "input": "output_1"
113 | }
114 | ]
115 | }
116 | },
117 | "outputs": {
118 | "output_1": {
119 | "connections": []
120 | }
121 | }
122 | }
123 | }
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/workstation/examples/6_reactagent.json:
--------------------------------------------------------------------------------
1 | {
2 | "2": {
3 | "name": "dashscope_chat",
4 | "data": {
5 | "args": {
6 | "config_name": "qwen",
7 | "model_name": "qwen-max",
8 | "api_key": "************",
9 | "temperature": 0.1,
10 | "seed": 1,
11 | "model_type": "dashscope_chat",
12 | "messages_key": "input"
13 | }
14 | },
15 | "inputs": {},
16 | "outputs": {}
17 | },
18 | "3": {
19 | "name": "ReActAgent",
20 | "data": {
21 | "elements": [
22 | "7",
23 | "5",
24 | "6",
25 | "8"
26 | ],
27 | "args": {
28 | "name": "Alice",
29 | "sys_prompt": "You are Alice",
30 | "model_config_name": "qwen",
31 | "max_iters": 10,
32 | "verbose": "True"
33 | }
34 | },
35 | "inputs": {
36 | "input_1": {
37 | "connections": [
38 | {
39 | "node": "4",
40 | "input": "output_1"
41 | }
42 | ]
43 | }
44 | },
45 | "outputs": {
46 | "output_1": {
47 | "connections": []
48 | }
49 | }
50 | },
51 | "4": {
52 | "name": "Message",
53 | "data": {
54 | "args": {
55 | "name": "Host",
56 | "content": "Hello",
57 | "url": ""
58 | }
59 | },
60 | "inputs": {
61 | "input_1": {
62 | "connections": []
63 | }
64 | },
65 | "outputs": {
66 | "output_1": {
67 | "connections": [
68 | {
69 | "node": "3",
70 | "output": "input_1"
71 | }
72 | ]
73 | }
74 | }
75 | },
76 | "5": {
77 | "name": "BingSearchService",
78 | "data": {
79 | "args": {
80 | "api_key": "************",
81 | "num_results": 3
82 | }
83 | },
84 | "inputs": {},
85 | "outputs": {}
86 | },
87 | "6": {
88 | "name": "PythonService",
89 | "data": {},
90 | "inputs": {},
91 | "outputs": {}
92 | },
93 | "7": {
94 | "name": "ReadTextService",
95 | "data": {},
96 | "inputs": {},
97 | "outputs": {}
98 | },
99 | "8": {
100 | "name": "WriteTextService",
101 | "data": {},
102 | "inputs": {},
103 | "outputs": {}
104 | }
105 | }
--------------------------------------------------------------------------------
/agentscope-main/src/agentscope/web/workstation/workflow_utils.py:
--------------------------------------------------------------------------------
1 | # -*- coding: utf-8 -*-
2 | """Workflow node utils."""
3 |
4 |
5 | def is_callable_expression(s: str) -> bool:
6 | """Check a expression whether a callable expression"""
7 | try:
8 | # Do not detect exp like this
9 | if s in ["input", "print"]:
10 | return False
11 | result = eval(s)
12 | return callable(result)
13 | except Exception:
14 | return False
15 |
16 |
17 | def kwarg_converter(kwargs: dict) -> str:
18 | """Convert a kwarg dict to a string."""
19 | kwarg_parts = []
20 | for key, value in kwargs.items():
21 | if is_callable_expression(value):
22 | kwarg_parts.append(f"{key}={value}")
23 | else:
24 | kwarg_parts.append(f"{key}={repr(value)}")
25 | return ", ".join(kwarg_parts)
26 |
27 |
28 | def deps_converter(dep_vars: list) -> str:
29 | """Convert a dep_vars list to a string."""
30 | return f"[{', '.join(dep_vars)}]"
31 |
32 |
33 | def dict_converter(dictionary: dict) -> str:
34 | """Convert a dictionary to a string."""
35 | result_parts = []
36 | for key, value in dictionary.items():
37 | result_parts.append(f'"{key}": {value}')
38 | return "{" + ", ".join(result_parts) + "}"
39 |
--------------------------------------------------------------------------------
/logo.png:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/open-sciencelab/Virtual-Scientists/9a6d0f32f9adbccef68f70f1705fa12fc9d10e23/logo.png
--------------------------------------------------------------------------------
/preprocess_data/CS_data/data_extraction/extract_coauthor2json.py:
--------------------------------------------------------------------------------
1 | import json
2 | import re
3 | from tqdm import tqdm
4 |
5 | def extract_coauthors(output_dir, filename):
6 | with open(filename, 'r') as file:
7 | data = file.read()
8 |
9 | # Split the data into blocks based on the author index
10 | coauthors_blocks = data.split('#')[1:] # Ignore the first empty split
11 |
12 | author2coauthor = {}
13 | for block in tqdm(coauthors_blocks):
14 | info_str = block.strip('\n').split('\t')
15 |
16 | author1 = info_str[0]
17 | author2 = info_str[1]
18 | weight = info_str[2]
19 | if author1 not in author2coauthor.keys():
20 | author2coauthor[author1] = []
21 | if author2 not in author2coauthor.keys():
22 | author2coauthor[author2] = []
23 |
24 | author2coauthor[author1].append({
25 | 'coauthor_id':author2,
26 | 'weight':weight
27 | })
28 |
29 | author2coauthor[author2].append({
30 | 'coauthor_id':author1,
31 | 'weight':weight
32 | })
33 |
34 | with open('{}/author2coauthor.json'.format(output_dir), 'w') as json_file:
35 | json.dump(author2coauthor, json_file, indent=4)
36 |
37 | if __name__=='__main__':
38 | # Example usage
39 | output_dir = '/home/bingxing2/ailab/group/ai4agr/shy/s4s'
40 | filename = 'AMiner-Coauthor.txt'
41 | extract_coauthors(output_dir, filename)
42 |
--------------------------------------------------------------------------------
/preprocess_data/CS_data/data_extraction/extract_publication2json.py:
--------------------------------------------------------------------------------
1 | import json
2 | import re
3 | from tqdm import tqdm
4 |
5 | def extract_papers_authors(output_dir, filename):
6 | with open(filename, 'r') as file:
7 | data = file.read()
8 |
9 | # Split the data into blocks based on the author index
10 | blocks = data.split('\n')[:-1] # Ignore the first empty split
11 |
12 | paper2author = {}
13 | author2paper = {}
14 | for block in tqdm(blocks):
15 | info_str = block.strip('\n').split('\t')
16 |
17 | author = info_str[1]
18 | paper = info_str[2]
19 |
20 | if author not in author2paper.keys():
21 | author2paper[author] = []
22 | if paper not in paper2author.keys():
23 | paper2author[paper] = []
24 |
25 | author2paper[author].append(paper)
26 | paper2author[paper].append(author)
27 |
28 | with open('{}/author2paper.json'.format(output_dir), 'w') as json_file:
29 | json.dump(author2paper, json_file, indent=4)
30 |
31 | with open('{}/paper2author.json'.format(output_dir), 'w') as json_file:
32 | json.dump(paper2author, json_file, indent=4)
33 |
34 | if __name__=='__main__':
35 | # Example usage
36 | output_dir = '/home/bingxing2/ailab/group/ai4agr/shy/s4s'
37 | filename = 'AMiner-Author2Paper.txt'
38 | extract_papers_authors(output_dir, filename)
39 |
--------------------------------------------------------------------------------
/preprocess_data/CS_data/database/add_paper_citation_num.py:
--------------------------------------------------------------------------------
1 | import sqlite3
2 | import os
3 | import json
4 | import re
5 | from tqdm import tqdm
6 | def add_paper_citation_num(root_dir):
7 | global_database_name = 'global_database.db'
8 | # connect to database, if it doesn't exist then build one
9 | conn = sqlite3.connect(os.path.join(root_dir, global_database_name))
10 | # create cursor
11 | cursor = conn.cursor()
12 |
13 | cursor.execute(f"SELECT * FROM papers")
14 | all_papers = cursor.fetchall()
15 |
16 | paper_citation = {}
17 | for paper in tqdm(all_papers):
18 | refs = paper[6].split(';') if paper[6] is not None else None
19 | if refs is None:
20 | continue
21 |
22 | for ref in refs:
23 | ref = ref.strip()
24 | if ref not in paper_citation.keys():
25 | paper_citation[ref] = 1
26 | else:
27 | paper_citation[ref] += 1
28 |
29 | cursor.execute('ALTER TABLE papers ADD COLUMN citation_num INTEGER')
30 | for k, v in tqdm(paper_citation.items()):
31 | cursor.execute(f'UPDATE papers SET citation_num = {v} WHERE id = {int(k)}')
32 |
33 | conn.commit()
34 | cursor.close()
35 | conn.close()
36 |
37 | if __name__=='__main__':
38 | # data dir
39 | root_dir = '/home/bingxing2/ailab/group/ai4agr/shy/s4s'
40 | add_paper_citation_num(root_dir)
--------------------------------------------------------------------------------
/preprocess_data/CS_data/visualization/author_affiliation_distribution.py:
--------------------------------------------------------------------------------
1 | import json
2 | from tqdm import tqdm
3 | import os
4 | import sqlite3
5 |
6 | replace_dict = {
7 | 'univ.':'university',
8 | 'dept.':'department'
9 | }
10 |
11 | def affiliation_str_clean(affiliation):
12 | for k, v in replace_dict.items():
13 | affiliation = affiliation.replace(k, v)
14 | affiliation = affiliation.replace(' ', ' ')
15 | return affiliation
16 |
17 | def extract_affiliations(root_dir, data_dir):
18 | # load database
19 | database_dir = os.path.join(root_dir, 'global_database.db')
20 | conn = sqlite3.connect(database_dir)
21 | cursor = conn.cursor()
22 |
23 | # load authors
24 | with open('{}/agentID2authorID.json'.format(data_dir), 'r') as file:
25 | agentID2authorID = json.load(file)
26 |
27 | # load affiliatioons
28 | affiliations = {}
29 | for agentID, authorID in tqdm(agentID2authorID.items()):
30 | cursor.execute(f'SELECT affiliations FROM authors WHERE id = {authorID}')
31 | rows = cursor.fetchall()
32 | affis = rows[0][0].split(';')
33 | for affi in affis:
34 | affi_split = affi.strip().split(',')
35 | affi1 = affiliation_str_clean(affi_split[0].strip())
36 | try:
37 | affiliations[affi1] += 1
38 | except:
39 | affiliations[affi1] = 1
40 | if len(affi_split)>=3:
41 | affi2 = affiliation_str_clean(affi_split[1].strip())
42 | try:
43 | affiliations[affi2] += 1
44 | except:
45 | affiliations[affi2] = 1
46 |
47 | return affiliations
48 |
49 | if __name__ == '__main__':
50 | # load the karate club graph
51 | root_dir = '/home/bingxing2/ailab/group/ai4agr/shy/s4s'
52 | used_data_dir = 'authors_degree_ge50_from_year2000to2010'
53 | affiliations = extract_affiliations(root_dir, os.path.join(root_dir, used_data_dir))
54 | affiliations = [[k, v] for k, v in affiliations.items()]
55 | affiliations = sorted(affiliations, key=lambda x:x[1], reverse=True)
56 |
57 | # save result
58 | save_dir = 'results'
59 | if not os.path.exists(save_dir):
60 | os.mkdir(save_dir)
61 | with open(f'{save_dir}/affiliation.txt', 'w') as file:
62 | file.write(f"there are {len(affiliations)} affiliations\n")
63 | for affi in affiliations:
64 | file.write(f"{affi[0]}:{affi[1]}\n")
--------------------------------------------------------------------------------
/preprocess_data/CS_data/visualization/author_citation_distribution.py:
--------------------------------------------------------------------------------
1 | import json
2 | from tqdm import tqdm
3 | import os
4 | import sqlite3
5 | import matplotlib.pyplot as plt
6 | import pandas as pd
7 |
8 | def extract_citations(root_dir, used_data_dir):
9 | # load database
10 | database_dir = os.path.join(root_dir, 'database.db')
11 | conn = sqlite3.connect(database_dir)
12 | cursor = conn.cursor()
13 |
14 | # load authors
15 | with open('{}/agentID2authorID.json'.format(os.path.join(root_dir, used_data_dir)), 'r') as file:
16 | agentID2authorID = json.load(file)
17 |
18 | # load affiliatioons
19 | citations = []
20 | for agentID, authorID in tqdm(agentID2authorID.items()):
21 | cursor.execute(f'SELECT citation_num FROM users WHERE id = {authorID}')
22 | rows = cursor.fetchall()
23 | citations.append(rows[0][0])
24 | return citations
25 |
26 | def draw_citation_histogram(citations):
27 | # Create a DataFrame
28 | df = pd.DataFrame(citations, columns=['Citations'])
29 |
30 | # Plot the histogram
31 | plt.figure(figsize=(10, 6))
32 | plt.hist(df['Citations'], bins=range(min(citations), max(citations), 200), edgecolor='black')
33 |
34 | plt.xticks(range(0, 16000, 1000),
35 | labels=range(0, 160, 10))
36 |
37 | # Show the plot
38 | # Add titles and labels
39 | plt.title('Histogram of Authors by Citation Range')
40 | plt.xlabel('Citation Range (x100)')
41 | plt.ylabel('Number of Authors')
42 |
43 | save_dir = 'results'
44 | if not os.path.exists(save_dir):
45 | os.mkdir(save_dir)
46 | plt.savefig(f'{save_dir}/citation.png', bbox_inches='tight')
47 |
48 | if __name__ == '__main__':
49 | # load the karate club graph
50 | root_dir = '/home/bingxing2/ailab/group/ai4agr/shy/s4s'
51 | used_data_dir = '156authors_degree_ge50_from_year2000to2010'
52 | citations = extract_citations(root_dir, used_data_dir)
53 | draw_citation_histogram(citations)
--------------------------------------------------------------------------------
/preprocess_data/CS_data/visualization/author_topic_distrbution.py:
--------------------------------------------------------------------------------
1 | import json
2 | from tqdm import tqdm
3 | import os
4 | import sqlite3
5 |
6 | replace_dict = {
7 | 'univ.':'university',
8 | 'dept.':'department'
9 | }
10 |
11 | def affiliation_str_clean(affiliation):
12 | for k, v in replace_dict.items():
13 | affiliation = affiliation.replace(k, v)
14 | affiliation = affiliation.replace(' ', ' ')
15 | return affiliation
16 |
17 | def extract_topics(root_dir, used_data_dir):
18 | # load database
19 | database_dir = os.path.join(root_dir, 'global_database.db')
20 | conn = sqlite3.connect(database_dir)
21 | cursor = conn.cursor()
22 |
23 | # load authors
24 | with open('{}/agentID2authorID.json'.format(os.path.join(root_dir, used_data_dir)), 'r') as file:
25 | agentID2authorID = json.load(file)
26 |
27 | # load affiliatioons
28 | all_topics = {}
29 | for agentID, authorID in tqdm(agentID2authorID.items()):
30 | cursor.execute(f'SELECT topics FROM users WHERE id = {authorID}')
31 | rows = cursor.fetchall()
32 | topics = rows[0][0].split(';')
33 | for topic in topics:
34 | topic_splits = topic.strip().split(',')
35 | for topic_split in topic_splits:
36 | try:
37 | all_topics[topic_split.strip()] += 1
38 | except:
39 | all_topics[topic_split.strip()] = 1
40 |
41 | return all_topics
42 |
43 | if __name__ == '__main__':
44 | # load the karate club graph
45 | root_dir = '/home/bingxing2/ailab/group/ai4agr/shy/s4s'
46 | used_data_dir = '156authors_degree_ge50_from_year2000to2010'
47 | topics = extract_topics(root_dir, used_data_dir)
48 | topics = [[k, v] for k, v in topics.items()]
49 | topics = sorted(topics, key=lambda x:x[1], reverse=True)
50 |
51 | # save result
52 | save_dir = 'results'
53 | if not os.path.exists(save_dir):
54 | os.mkdir(save_dir)
55 | with open(f'{save_dir}/topic.txt', 'w') as file:
56 | file.write(f"there are {len(topics)} topics\n")
57 | for topic in topics:
58 | file.write(f"{topic[0]}:{topic[1]}\n")
--------------------------------------------------------------------------------
/sci_platform/.DS_Store:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/open-sciencelab/Virtual-Scientists/9a6d0f32f9adbccef68f70f1705fa12fc9d10e23/sci_platform/.DS_Store
--------------------------------------------------------------------------------
/sci_platform/__pycache__/platform.cpython-311.pyc:
--------------------------------------------------------------------------------
https://raw.githubusercontent.com/open-sciencelab/Virtual-Scientists/9a6d0f32f9adbccef68f70f1705fa12fc9d10e23/sci_platform/__pycache__/platform.cpython-311.pyc
--------------------------------------------------------------------------------
/sci_platform/configs/knowledge_config.json:
--------------------------------------------------------------------------------
1 | [
2 | {
3 | "knowledge_id": "professional paper",
4 | "emb_model_config_name": "ollama_embedding-mxbai-embed-large",
5 | "data_processing": [
6 | {
7 | "load_data": {
8 | "loader": {
9 | "create_object": true,
10 | "module": "llama_index.core",
11 | "class": "SimpleDirectoryReader",
12 | "init_args": {
13 | "input_dir": "/home/bingxing2/ailab/group/ai4agr/crq/SciSci/books",
14 | "required_exts": [
15 | ".txt"
16 | ]
17 | }
18 | }
19 | }
20 | }
21 | ]
22 | }
23 | ]
--------------------------------------------------------------------------------
/sci_platform/configs/model_configs.json:
--------------------------------------------------------------------------------
1 | [
2 | {
3 | "config_name": "ollama_llama3.1_8b",
4 | "model_type": "ollama_chat",
5 | "model_name": "llama3.1",
6 | "options": {
7 | "temperature": 0.5,
8 | "seed": 123
9 | },
10 | "keep_alive": "5m"
11 | },
12 | {
13 | "config_name": "ollama_llama3.1_70b",
14 | "model_type": "ollama_chat",
15 | "model_name": "llama3.1:70b",
16 | "options": {
17 | "temperature": 0.5,
18 | "seed": 123
19 | },
20 | "keep_alive": "5m"
21 | },
22 | {
23 | "config_name": "ollama_llama3_8b",
24 | "model_type": "ollama_chat",
25 | "model_name": "llama3",
26 | "options": {
27 | "temperature": 0.5,
28 | "seed": 123
29 | },
30 | "keep_alive": "5m"
31 | },
32 | {
33 | "config_name": "ollama_llama3_70b",
34 | "model_type": "ollama_chat",
35 | "model_name": "llama3:70b",
36 | "options": {
37 | "temperature": 0.5,
38 | "seed": 123
39 | },
40 | "keep_alive": "5m"
41 | },
42 | {
43 | "config_name": "ollama_embedding-qwen:0.5b",
44 | "model_type": "ollama_embedding",
45 | "model_name": "qwen:0.5b",
46 | "options": {
47 | "temperature": 0.7
48 | },
49 | "keep_alive": "5m"
50 | },
51 | {
52 | "config_name": "ollama_embedding-mxbai-embed-large",
53 | "model_type": "ollama_embedding",
54 | "model_name": "mxbai-embed-large",
55 | "options": {
56 | "temperature": 0.7
57 | },
58 | "keep_alive": "5m"
59 | }
60 | ]
--------------------------------------------------------------------------------
/sci_platform/run.py:
--------------------------------------------------------------------------------
1 | from sci_platform import Platform
2 | from utils.scientist_utils import read_txt_files_as_dict
3 | import os
4 | import argparse
5 | def parse_arguments():
6 | parser = argparse.ArgumentParser(description="Run experiments")
7 | # parser.add_argument(
8 | # "--skip-idea-generation",
9 | # action="store_true",
10 | # help="Skip idea generation and load existing ideas",
11 | # )
12 | # parser.add_argument(
13 | # "--skip-novelty-check",
14 | # action="store_true",
15 | # help="Skip novelty check and use existing ideas",
16 | # )
17 |
18 | # how many runs
19 | parser.add_argument(
20 | "--runs",
21 | type=int,
22 | default=10,
23 | help="Calculate average on how many runs.",
24 | )
25 | # team limit
26 | parser.add_argument(
27 | "--team_limit",
28 | type=int,
29 | default=2,
30 | help="Max number of teams for a scientist.",
31 | )
32 | parser.add_argument(
33 | "--max_discuss_iteration",
34 | type=int,
35 | default=2,
36 | help="Max discuss iteration.",
37 | )
38 | parser.add_argument(
39 | "--max_team_member",
40 | type=int,
41 | default=2,
42 | help="Max team mamber of a team, actual team size is max_team_member.",
43 | )
44 | parser.add_argument(
45 | "--epochs",
46 | type=int,
47 | default=6,
48 | help="Epochs.",
49 | )
50 |
51 | return parser.parse_args()
52 |
53 | if __name__ == '__main__':
54 | args = parse_arguments()
55 |
56 | args.save_dir = f'team_info/{args.max_discuss_iteration}_itrs_{args.max_team_member}_members'
57 | args.log_dir = f'team_log/{args.max_discuss_iteration}_itrs_{args.max_team_member}_members'
58 |
59 | end =False
60 | os.makedirs(args.save_dir, exist_ok=True)
61 | os.makedirs(args.log_dir, exist_ok=True)
62 |
63 | while end==False:
64 | print(f'{len(os.listdir(args.save_dir))} files are created...')
65 | platform_example = Platform(
66 | team_limit = args.team_limit,
67 | group_max_discuss_iteration = args.max_discuss_iteration,
68 | max_teammember = args.max_team_member-1,
69 | log_dir = args.log_dir,
70 | info_dir = args.save_dir
71 | )
72 | platform_example.running(args.epochs)
73 | if len(os.listdir(args.save_dir)) >= args.team_limit*args.runs:
74 | end = True
75 |
--------------------------------------------------------------------------------
/sci_platform/sci_team/__init__.py:
--------------------------------------------------------------------------------
1 | from .SciTeam import Team
--------------------------------------------------------------------------------
/sci_platform/utils/__init__.py:
--------------------------------------------------------------------------------
1 | from .prompt import Prompts
2 | from .scientist_utils import *
--------------------------------------------------------------------------------
18 | 
23 |