yujunjun/TradingAgents
1
0
Fork 0
mirror of https://github.com/TauricResearch/TradingAgents.git synced 2025-06-27 00:41:34 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'TauricResearch:main' into save_results

Browse Source
This commit is contained in:
Huijae Lee 2025-06-25 08:43:19 +09:00 committed by GitHub
commit ee3d499894
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
17 changed files with 5674 additions and 56 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
.python-version Normal file
View File

@ -0,0 +1 @@
3.10

4
README.md
View File

@ -80,7 +80,7 @@ Our framework decomposes complex trading tasks into specialized roles. This ensu
- Composes reports from the analysts and researchers to make informed trading decisions. It determines the timing and magnitude of trades based on comprehensive market insights.
<p align="center">
<img src="assets/risk.png" width="70%" style="display: inline-block; margin: 0 2%;">
<img src="assets/trader.png" width="70%" style="display: inline-block; margin: 0 2%;">
</p>
### Risk Management and Portfolio Manager
@ -88,7 +88,7 @@ Our framework decomposes complex trading tasks into specialized roles. This ensu
- The Portfolio Manager approves/rejects the transaction proposal. If approved, the order will be sent to the simulated exchange and executed.
<p align="center">
<img src="assets/trader.png" width="70%" style="display: inline-block; margin: 0 2%;">
<img src="assets/risk.png" width="70%" style="display: inline-block; margin: 0 2%;">
</p>
## Installation and CLI

63
cli/main.py
View File

@ -297,10 +297,27 @@ def update_display(layout, spinner_text=None):
# Add regular messages
for timestamp, msg_type, content in message_buffer.messages:
# Convert content to string if it's not already
content_str = content
if isinstance(content, list):
# Handle list of content blocks (Anthropic format)
text_parts = []
for item in content:
if isinstance(item, dict):
if item.get('type') == 'text':
text_parts.append(item.get('text', ''))
elif item.get('type') == 'tool_use':
text_parts.append(f"[Tool: {item.get('name', 'unknown')}]")
else:
text_parts.append(str(item))
content_str = ' '.join(text_parts)
elif not isinstance(content_str, str):
content_str = str(content)
# Truncate message content if too long
if isinstance(content, str) and len(content) > 200:
content = content[:197] + "..."
all_messages.append((timestamp, msg_type, content))
if len(content_str) > 200:
content_str = content_str[:197] + "..."
all_messages.append((timestamp, msg_type, content_str))
# Sort by timestamp
all_messages.sort(key=lambda x: x[0])
@ -446,20 +463,30 @@ def get_user_selections():
)
selected_research_depth = select_research_depth()
# Step 5: Thinking agents
# Step 5: OpenAI backend
console.print(
create_question_box(
"Step 5: Thinking Agents", "Select your thinking agents for analysis"
"Step 5: OpenAI backend", "Select which service to talk to"
)
)
selected_shallow_thinker = select_shallow_thinking_agent()
selected_deep_thinker = select_deep_thinking_agent()
selected_llm_provider, backend_url = select_llm_provider()
# Step 6: Thinking agents
console.print(
create_question_box(
"Step 6: Thinking Agents", "Select your thinking agents for analysis"
)
)
selected_shallow_thinker = select_shallow_thinking_agent(selected_llm_provider)
selected_deep_thinker = select_deep_thinking_agent(selected_llm_provider)
return {
"ticker": selected_ticker,
"analysis_date": analysis_date,
"analysts": selected_analysts,
"research_depth": selected_research_depth,
"llm_provider": selected_llm_provider.lower(),
"backend_url": backend_url,
"shallow_thinker": selected_shallow_thinker,
"deep_thinker": selected_deep_thinker,
}
@ -685,6 +712,24 @@ def update_research_team_status(status):
for agent in research_team:
message_buffer.update_agent_status(agent, status)
def extract_content_string(content):
"""Extract string content from various message formats."""
if isinstance(content, str):
return content
elif isinstance(content, list):
# Handle Anthropic's list format
text_parts = []
for item in content:
if isinstance(item, dict):
if item.get('type') == 'text':
text_parts.append(item.get('text', ''))
elif item.get('type') == 'tool_use':
text_parts.append(f"[Tool: {item.get('name', 'unknown')}]")
else:
text_parts.append(str(item))
return ' '.join(text_parts)
else:
return str(content)
def run_analysis():
# First get all user selections
@ -696,6 +741,8 @@ def run_analysis():
config["max_risk_discuss_rounds"] = selections["research_depth"]
config["quick_think_llm"] = selections["shallow_thinker"]
config["deep_think_llm"] = selections["deep_thinker"]
config["backend_url"] = selections["backend_url"]
config["llm_provider"] = selections["llm_provider"].lower()
# Initialize the graph
graph = TradingAgentsGraph(
@ -803,7 +850,7 @@ def run_analysis():
# Extract message content and type
if hasattr(last_message, "content"):
content = last_message.content
content = extract_content_string(last_message.content) # Use the helper function
msg_type = "Reasoning"
else:
content = str(last_message)

91
cli/utils.py
View File

@ -122,22 +122,43 @@ def select_research_depth() -> int:
return choice
def select_shallow_thinking_agent() -> str:
def select_shallow_thinking_agent(provider) -> str:
"""Select shallow thinking llm engine using an interactive selection."""
# Define shallow thinking llm engine options with their corresponding model names
SHALLOW_AGENT_OPTIONS = [
SHALLOW_AGENT_OPTIONS = {
"openai": [
("GPT-4o-mini - Fast and efficient for quick tasks", "gpt-4o-mini"),
("GPT-4.1-nano - Ultra-lightweight model for basic operations", "gpt-4.1-nano"),
("GPT-4.1-mini - Compact model with good performance", "gpt-4.1-mini"),
("GPT-4o - Standard model with solid capabilities", "gpt-4o"),
],
"anthropic": [
("Claude Haiku 3.5 - Fast inference and standard capabilities", "claude-3-5-haiku-latest"),
("Claude Sonnet 3.5 - Highly capable standard model", "claude-3-5-sonnet-latest"),
("Claude Sonnet 3.7 - Exceptional hybrid reasoning and agentic capabilities", "claude-3-7-sonnet-latest"),
("Claude Sonnet 4 - High performance and excellent reasoning", "claude-sonnet-4-0"),
],
"google": [
("Gemini 2.0 Flash-Lite - Cost efficiency and low latency", "gemini-2.0-flash-lite"),
("Gemini 2.0 Flash - Next generation features, speed, and thinking", "gemini-2.0-flash"),
("Gemini 2.5 Flash - Adaptive thinking, cost efficiency", "gemini-2.5-flash-preview-05-20"),
],
"openrouter": [
("Meta: Llama 4 Scout", "meta-llama/llama-4-scout:free"),
("Meta: Llama 3.3 8B Instruct - A lightweight and ultra-fast variant of Llama 3.3 70B", "meta-llama/llama-3.3-8b-instruct:free"),
("google/gemini-2.0-flash-exp:free - Gemini Flash 2.0 offers a significantly faster time to first token", "google/gemini-2.0-flash-exp:free"),
],
"ollama": [
("llama3.2 local", "llama3.2"),
]
}
choice = questionary.select(
"Select Your [Quick-Thinking LLM Engine]:",
choices=[
questionary.Choice(display, value=value)
for display, value in SHALLOW_AGENT_OPTIONS
for display, value in SHALLOW_AGENT_OPTIONS[provider.lower()]
],
instruction="\n- Use arrow keys to navigate\n- Press Enter to select",
style=questionary.Style(
@ -158,11 +179,12 @@ def select_shallow_thinking_agent() -> str:
return choice
def select_deep_thinking_agent() -> str:
def select_deep_thinking_agent(provider) -> str:
"""Select deep thinking llm engine using an interactive selection."""
# Define deep thinking llm engine options with their corresponding model names
DEEP_AGENT_OPTIONS = [
DEEP_AGENT_OPTIONS = {
"openai": [
("GPT-4.1-nano - Ultra-lightweight model for basic operations", "gpt-4.1-nano"),
("GPT-4.1-mini - Compact model with good performance", "gpt-4.1-mini"),
("GPT-4o - Standard model with solid capabilities", "gpt-4o"),
@ -170,13 +192,34 @@ def select_deep_thinking_agent() -> str:
("o3-mini - Advanced reasoning model (lightweight)", "o3-mini"),
("o3 - Full advanced reasoning model", "o3"),
("o1 - Premier reasoning and problem-solving model", "o1"),
],
"anthropic": [
("Claude Haiku 3.5 - Fast inference and standard capabilities", "claude-3-5-haiku-latest"),
("Claude Sonnet 3.5 - Highly capable standard model", "claude-3-5-sonnet-latest"),
("Claude Sonnet 3.7 - Exceptional hybrid reasoning and agentic capabilities", "claude-3-7-sonnet-latest"),
("Claude Sonnet 4 - High performance and excellent reasoning", "claude-sonnet-4-0"),
("Claude Opus 4 - Most powerful Anthropic model", " claude-opus-4-0"),
],
"google": [
("Gemini 2.0 Flash-Lite - Cost efficiency and low latency", "gemini-2.0-flash-lite"),
("Gemini 2.0 Flash - Next generation features, speed, and thinking", "gemini-2.0-flash"),
("Gemini 2.5 Flash - Adaptive thinking, cost efficiency", "gemini-2.5-flash-preview-05-20"),
("Gemini 2.5 Pro", "gemini-2.5-pro-preview-06-05"),
],
"openrouter": [
("DeepSeek V3 - a 685B-parameter, mixture-of-experts model", "deepseek/deepseek-chat-v3-0324:free"),
("Deepseek - latest iteration of the flagship chat model family from the DeepSeek team.", "deepseek/deepseek-chat-v3-0324:free"),
],
"ollama": [
("qwen3", "qwen3"),
]
}
choice = questionary.select(
"Select Your [Deep-Thinking LLM Engine]:",
choices=[
questionary.Choice(display, value=value)
for display, value in DEEP_AGENT_OPTIONS
for display, value in DEEP_AGENT_OPTIONS[provider.lower()]
],
instruction="\n- Use arrow keys to navigate\n- Press Enter to select",
style=questionary.Style(
@ -193,3 +236,39 @@ def select_deep_thinking_agent() -> str:
exit(1)
return choice
def select_llm_provider() -> tuple[str, str]:
"""Select the OpenAI api url using interactive selection."""
# Define OpenAI api options with their corresponding endpoints
BASE_URLS = [
("OpenAI", "https://api.openai.com/v1"),
("Anthropic", "https://api.anthropic.com/"),
("Google", "https://generativelanguage.googleapis.com/v1"),
("Openrouter", "https://openrouter.ai/api/v1"),
("Ollama", "http://localhost:11434/v1"),
]
choice = questionary.select(
"Select your LLM Provider:",
choices=[
questionary.Choice(display, value=(display, value))
for display, value in BASE_URLS
],
instruction="\n- Use arrow keys to navigate\n- Press Enter to select",
style=questionary.Style(
[
("selected", "fg:magenta noinherit"),
("highlighted", "fg:magenta noinherit"),
("pointer", "fg:magenta noinherit"),
]
),
).ask()
if choice is None:
console.print("\n[red]no OpenAI backend selected. Exiting...[/red]")
exit(1)
display_name, url = choice
print(f"You selected: {display_name}\tURL: {url}")
return display_name, url

6
main.py
View File

@ -3,8 +3,10 @@ from tradingagents.default_config import DEFAULT_CONFIG
# Create a custom config
config = DEFAULT_CONFIG.copy()
config["deep_think_llm"] = "gpt-4.1-nano" # Use a different model
config["quick_think_llm"] = "gpt-4.1-nano" # Use a different model
config["llm_provider"] = "google" # Use a different model
config["backend_url"] = "https://generativelanguage.googleapis.com/v1" # Use a different backend
config["deep_think_llm"] = "gemini-2.0-flash" # Use a different model
config["quick_think_llm"] = "gemini-2.0-flash" # Use a different model
config["max_debate_rounds"] = 1 # Increase debate rounds
config["online_tools"] = True # Increase debate rounds

34
pyproject.toml Normal file
View File

@ -0,0 +1,34 @@
[project]
name = "tradingagents"
version = "0.1.0"
description = "Add your description here"
readme = "README.md"
requires-python = ">=3.10"
dependencies = [
"akshare>=1.16.98",
"backtrader>=1.9.78.123",
"chainlit>=2.5.5",
"chromadb>=1.0.12",
"eodhd>=1.0.32",
"feedparser>=6.0.11",
"finnhub-python>=2.4.23",
"langchain-anthropic>=0.3.15",
"langchain-experimental>=0.3.4",
"langchain-google-genai>=2.1.5",
"langchain-openai>=0.3.23",
"langgraph>=0.4.8",
"pandas>=2.3.0",
"parsel>=1.10.0",
"praw>=7.8.1",
"pytz>=2025.2",
"questionary>=2.1.0",
"redis>=6.2.0",
"requests>=2.32.4",
"rich>=14.0.0",
"setuptools>=80.9.0",
"stockstats>=0.6.5",
"tqdm>=4.67.1",
"tushare>=1.4.21",
"typing-extensions>=4.14.0",
"yfinance>=0.2.63",
]

2
requirements.txt
View File

@ -22,3 +22,5 @@ redis
chainlit
rich
questionary
langchain_anthropic
langchain-google-genai

9
tradingagents/agents/analysts/fundamentals_analyst.py
View File

@ -22,7 +22,7 @@ def create_fundamentals_analyst(llm, toolkit):
system_message = (
"You are a researcher tasked with analyzing fundamental information over the past week about a company. Please write a comprehensive report of the company's fundamental information such as financial documents, company profile, basic company financials, company financial history, insider sentiment and insider transactions to gain a full view of the company's fundamental information to inform traders. Make sure to include as much detail as possible. Do not simply state the trends are mixed, provide detailed and finegrained analysis and insights that may help traders make decisions."
+ " Make sure to append a Makrdown table at the end of the report to organize key points in the report, organized and easy to read.",
+ " Make sure to append a Markdown table at the end of the report to organize key points in the report, organized and easy to read.",
)
prompt = ChatPromptTemplate.from_messages(
@ -51,9 +51,14 @@ def create_fundamentals_analyst(llm, toolkit):
result = chain.invoke(state["messages"])
report = ""
if len(result.tool_calls) == 0:
report = result.content
return {
"messages": [result],
"fundamentals_report": result.content,
"fundamentals_report": report,
}
return fundamentals_analyst_node

7
tradingagents/agents/analysts/market_analyst.py
View File

@ -76,9 +76,14 @@ Volume-Based Indicators:
result = chain.invoke(state["messages"])
report = ""
if len(result.tool_calls) == 0:
report = result.content
return {
"messages": [result],
"market_report": result.content,
"market_report": report,
}
return market_analyst_node

7
tradingagents/agents/analysts/news_analyst.py
View File

@ -47,9 +47,14 @@ def create_news_analyst(llm, toolkit):
chain = prompt | llm.bind_tools(tools)
result = chain.invoke(state["messages"])
report = ""
if len(result.tool_calls) == 0:
report = result.content
return {
"messages": [result],
"news_report": result.content,
"news_report": report,
}
return news_analyst_node

7
tradingagents/agents/analysts/social_media_analyst.py
View File

@ -47,9 +47,14 @@ def create_social_media_analyst(llm, toolkit):
result = chain.invoke(state["messages"])
report = ""
if len(result.tool_calls) == 0:
report = result.content
return {
"messages": [result],
"sentiment_report": result.content,
"sentiment_report": report,
}
return social_media_analyst_node

12
tradingagents/agents/utils/agent_utils.py
View File

@ -12,13 +12,21 @@ from dateutil.relativedelta import relativedelta
from langchain_openai import ChatOpenAI
import tradingagents.dataflows.interface as interface
from tradingagents.default_config import DEFAULT_CONFIG
from langchain_core.messages import HumanMessage
def create_msg_delete():
def delete_messages(state):
"""To prevent message history from overflowing, regularly clear message history after a stage of the pipeline is done"""
"""Clear messages and add placeholder for Anthropic compatibility"""
messages = state["messages"]
return {"messages": [RemoveMessage(id=m.id) for m in messages]}
# Remove all messages
removal_operations = [RemoveMessage(id=m.id) for m in messages]
# Add a minimal placeholder message
placeholder = HumanMessage(content="Continue")
return {"messages": removal_operations + [placeholder]}
return delete_messages

10
tradingagents/agents/utils/memory.py
View File

@ -1,19 +1,23 @@
import chromadb
from chromadb.config import Settings
from openai import OpenAI
import numpy as np
class FinancialSituationMemory:
def __init__(self, name):
def __init__(self, name, config):
if config["backend_url"] == "http://localhost:11434/v1":
self.embedding = "nomic-embed-text"
else:
self.embedding = "text-embedding-3-small"
self.client = OpenAI()
self.chroma_client = chromadb.Client(Settings(allow_reset=True))
self.situation_collection = self.chroma_client.create_collection(name=name)
def get_embedding(self, text):
"""Get OpenAI embedding for a text"""
response = self.client.embeddings.create(
model="text-embedding-ada-002", input=text
model=self.embedding, input=text
)
return response.data[0].embedding

5
tradingagents/dataflows/interface.py
View File

@ -703,6 +703,7 @@ def get_YFin_data(
def get_stock_news_openai(ticker, curr_date):
config = get_config()
client = OpenAI()
response = client.responses.create(
@ -713,7 +714,7 @@ def get_stock_news_openai(ticker, curr_date):
"content": [
{
"type": "input_text",
"text": f"Can you search Social Media for {ticker} on TSLA from 7 days before {curr_date} to {curr_date}? Make sure you only get the data posted during that period.",
"text": f"Can you search Social Media for {ticker} from 7 days before {curr_date} to {curr_date}? Make sure you only get the data posted during that period.",
}
],
}
@ -737,6 +738,7 @@ def get_stock_news_openai(ticker, curr_date):
def get_global_news_openai(curr_date):
config = get_config()
client = OpenAI()
response = client.responses.create(
@ -771,6 +773,7 @@ def get_global_news_openai(curr_date):
def get_fundamentals_openai(ticker, curr_date):
config = get_config()
client = OpenAI()
response = client.responses.create(

2
tradingagents/default_config.py
View File

@ -9,8 +9,10 @@ DEFAULT_CONFIG = {
"dataflows/data_cache",
),
# LLM settings
"llm_provider": "openai",
"deep_think_llm": "o4-mini",
"quick_think_llm": "gpt-4o-mini",
"backend_url": "https://api.openai.com/v1",
# Debate and discussion settings
"max_debate_rounds": 1,
"max_risk_discuss_rounds": 1,

29
tradingagents/graph/trading_graph.py
View File

@ -7,6 +7,9 @@ from datetime import date
from typing import Dict, Any, Tuple, List, Optional
from langchain_openai import ChatOpenAI
from langchain_anthropic import ChatAnthropic
from langchain_google_genai import ChatGoogleGenerativeAI
from langgraph.prebuilt import ToolNode
from tradingagents.agents import *
@ -55,18 +58,26 @@ class TradingAgentsGraph:
)
# Initialize LLMs
self.deep_thinking_llm = ChatOpenAI(model=self.config["deep_think_llm"])
self.quick_thinking_llm = ChatOpenAI(
model=self.config["quick_think_llm"], temperature=0.1
)
if self.config["llm_provider"].lower() == "openai" or self.config["llm_provider"] == "ollama" or self.config["llm_provider"] == "openrouter":
self.deep_thinking_llm = ChatOpenAI(model=self.config["deep_think_llm"], base_url=self.config["backend_url"])
self.quick_thinking_llm = ChatOpenAI(model=self.config["quick_think_llm"], base_url=self.config["backend_url"])
elif self.config["llm_provider"].lower() == "anthropic":
self.deep_thinking_llm = ChatAnthropic(model=self.config["deep_think_llm"], base_url=self.config["backend_url"])
self.quick_thinking_llm = ChatAnthropic(model=self.config["quick_think_llm"], base_url=self.config["backend_url"])
elif self.config["llm_provider"].lower() == "google":
self.deep_thinking_llm = ChatGoogleGenerativeAI(model=self.config["deep_think_llm"])
self.quick_thinking_llm = ChatGoogleGenerativeAI(model=self.config["quick_think_llm"])
else:
raise ValueError(f"Unsupported LLM provider: {self.config['llm_provider']}")
self.toolkit = Toolkit(config=self.config)
# Initialize memories
self.bull_memory = FinancialSituationMemory("bull_memory")
self.bear_memory = FinancialSituationMemory("bear_memory")
self.trader_memory = FinancialSituationMemory("trader_memory")
self.invest_judge_memory = FinancialSituationMemory("invest_judge_memory")
self.risk_manager_memory = FinancialSituationMemory("risk_manager_memory")
self.bull_memory = FinancialSituationMemory("bull_memory", self.config)
self.bear_memory = FinancialSituationMemory("bear_memory", self.config)
self.trader_memory = FinancialSituationMemory("trader_memory", self.config)
self.invest_judge_memory = FinancialSituationMemory("invest_judge_memory", self.config)
self.risk_manager_memory = FinancialSituationMemory("risk_manager_memory", self.config)
# Create tool nodes
self.tool_nodes = self._create_tool_nodes()

5405
uv.lock generated Normal file
View File

File diff suppressed because it is too large Load Diff