charieswei
/
TradingAgents
mirror of https://github.com/TauricResearch/TradingAgents.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Updated to support news and fundamentals analysis for non-openai llms

This commit is contained in:
sdk451 2025-07-23 11:24:42 +10:00
parent b256159d9e
commit b600d59e31
3 changed files with 201 additions and 95 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
cli/main.py
View File

@ -771,7 +771,7 @@ def run_analysis():
func(*args, **kwargs) func(*args, **kwargs)
timestamp, message_type, content = obj.messages[-1] timestamp, message_type, content = obj.messages[-1]
content = content.replace("\n", " ") # Replace newlines with spaces content = content.replace("\n", " ") # Replace newlines with spaces
with open(log_file, "a") as f: with open(log_file, "a", encoding="utf-8") as f:
f.write(f"{timestamp} [{message_type}] {content}\n") f.write(f"{timestamp} [{message_type}] {content}\n")
return wrapper return wrapper
@ -782,7 +782,7 @@ def run_analysis():
func(*args, **kwargs) func(*args, **kwargs)
timestamp, tool_name, args = obj.tool_calls[-1] timestamp, tool_name, args = obj.tool_calls[-1]
args_str = ", ".join(f"{k}={v}" for k, v in args.items()) args_str = ", ".join(f"{k}={v}" for k, v in args.items())
with open(log_file, "a") as f: with open(log_file, "a", encoding="utf-8") as f:
f.write(f"{timestamp} [Tool Call] {tool_name}({args_str})\n") f.write(f"{timestamp} [Tool Call] {tool_name}({args_str})\n")
return wrapper return wrapper
@ -795,7 +795,7 @@ def run_analysis():
content = obj.report_sections[section_name] content = obj.report_sections[section_name]
if content: if content:
file_name = f"{section_name}.md" file_name = f"{section_name}.md"
with open(report_dir / file_name, "w") as f: with open(report_dir / file_name, "w", encoding="utf-8") as f:
f.write(content) f.write(content)
return wrapper return wrapper

285
tradingagents/dataflows/interface.py
View File

@ -14,6 +14,8 @@ from tqdm import tqdm
import yfinance as yf import yfinance as yf
from openai import OpenAI from openai import OpenAI
from .config import get_config, set_config, DATA_DIR from .config import get_config, set_config, DATA_DIR
from langchain_google_genai import ChatGoogleGenerativeAI
from langchain_anthropic import ChatAnthropic
def get_finnhub_news( def get_finnhub_news(
@ -704,104 +706,207 @@ def get_YFin_data(
def get_stock_news_openai(ticker, curr_date): def get_stock_news_openai(ticker, curr_date):
config = get_config() config = get_config()
client = OpenAI(base_url=config["backend_url"]) provider = config["llm_provider"].lower()
if provider == "openai":
client = OpenAI(base_url=config["backend_url"])
response = client.responses.create(
model=config["quick_think_llm"],
input=[
{
"role": "system",
"content": [
{
"type": "input_text",
"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.",
}
],
}
],
text={"format": {"type": "text"}},
reasoning={},
tools=[
{
"type": "web_search_preview",
"user_location": {"type": "approximate"},
"search_context_size": "low",
}
],
temperature=1,
max_output_tokens=4096,
top_p=1,
store=True,
)
return response.output[1].content[0].text
else:
# Use Finnhub and Google News/Reddit for real news/social data
news = get_finnhub_news(ticker, curr_date, 7)
reddit = get_reddit_company_news(ticker, curr_date, 7, 5)
google_news = get_google_news(ticker, curr_date, 7)
context = f"""
# Finnhub News
{news}
response = client.responses.create( # Reddit News
model=config["quick_think_llm"], {reddit}
input=[
{
"role": "system",
"content": [
{
"type": "input_text",
"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.",
}
],
}
],
text={"format": {"type": "text"}},
reasoning={},
tools=[
{
"type": "web_search_preview",
"user_location": {"type": "approximate"},
"search_context_size": "low",
}
],
temperature=1,
max_output_tokens=4096,
top_p=1,
store=True,
)
return response.output[1].content[0].text
# Google News
{google_news}
"""
prompt = (
f"Given the following real company news and social data for {ticker} as of {curr_date}, "
"write a detailed news and sentiment analysis. Summarize key events, sentiment, and any trends that would affect trading. "
"Present the information in a markdown table at the end.\n\n"
f"{context}"
)
if provider == "google":
google_api_key = os.getenv("GOOGLE_API_KEY")
llm = ChatGoogleGenerativeAI(model=config["quick_think_llm"], google_api_key=google_api_key)
result = llm.invoke(prompt)
return result.content
elif provider == "anthropic":
anthropic_api_key = os.getenv("ANTHROPIC_API_KEY")
llm = ChatAnthropic(model_name=config["quick_think_llm"], api_key=anthropic_api_key)
result = llm.invoke(prompt)
return result.content
else:
return "Stock news retrieval is not supported for this LLM provider."
def get_global_news_openai(curr_date): def get_global_news_openai(curr_date):
config = get_config() config = get_config()
client = OpenAI(base_url=config["backend_url"]) provider = config["llm_provider"].lower()
if provider == "openai":
response = client.responses.create( client = OpenAI(base_url=config["backend_url"])
model=config["quick_think_llm"], response = client.responses.create(
input=[ model=config["quick_think_llm"],
{ input=[
"role": "system", {
"content": [ "role": "system",
{ "content": [
"type": "input_text", {
"text": f"Can you search global or macroeconomics news from 7 days before {curr_date} to {curr_date} that would be informative for trading purposes? Make sure you only get the data posted during that period.", "type": "input_text",
} "text": f"Can you search global or macroeconomics news from 7 days before {curr_date} to {curr_date} that would be informative for trading purposes? Make sure you only get the data posted during that period.",
], }
} ],
], }
text={"format": {"type": "text"}}, ],
reasoning={}, text={"format": {"type": "text"}},
tools=[ reasoning={},
{ tools=[
"type": "web_search_preview", {
"user_location": {"type": "approximate"}, "type": "web_search_preview",
"search_context_size": "low", "user_location": {"type": "approximate"},
} "search_context_size": "low",
], }
temperature=1, ],
max_output_tokens=4096, temperature=1,
top_p=1, max_output_tokens=4096,
store=True, top_p=1,
) store=True,
)
return response.output[1].content[0].text return response.output[1].content[0].text
else:
reddit = get_reddit_global_news(curr_date, 7, 5)
google_news = get_google_news("global macroeconomics", curr_date, 7)
context = f"""
# Reddit Global News
{reddit}
# Google News
{google_news}
"""
prompt = (
f"Given the following real global and macroeconomic news as of {curr_date}, "
"write a detailed macroeconomic news analysis. Summarize key events, sentiment, and any trends that would affect trading. "
"Present the information in a markdown table at the end.\n\n"
f"{context}"
)
if provider == "google":
google_api_key = os.getenv("GOOGLE_API_KEY")
llm = ChatGoogleGenerativeAI(model=config["quick_think_llm"], google_api_key=google_api_key)
result = llm.invoke(prompt)
return result.content
elif provider == "anthropic":
anthropic_api_key = os.getenv("ANTHROPIC_API_KEY")
llm = ChatAnthropic(model_name=config["quick_think_llm"], api_key=anthropic_api_key)
result = llm.invoke(prompt)
return result.content
else:
return "Global news retrieval is not supported for this LLM provider."
def get_fundamentals_openai(ticker, curr_date): def get_fundamentals_openai(ticker, curr_date):
config = get_config() config = get_config()
client = OpenAI(base_url=config["backend_url"]) provider = config["llm_provider"].lower()
if provider == "openai":
client = OpenAI(base_url=config["backend_url"])
response = client.responses.create(
model=config["quick_think_llm"],
input=[
{
"role": "system",
"content": [
{
"type": "input_text",
"text": f"Can you search Fundamental for discussions on {ticker} during of the month before {curr_date} to the month of {curr_date}. Make sure you only get the data posted during that period. List as a table, with PE/PS/Cash flow/ etc",
}
],
}
],
text={"format": {"type": "text"}},
reasoning={},
tools=[
{
"type": "web_search_preview",
"user_location": {"type": "approximate"},
"search_context_size": "low",
}
],
temperature=1,
max_output_tokens=4096,
top_p=1,
store=True,
)
return response.output[1].content[0].text
else:
# Fetch fundamentals from Yahoo Finance
import yfinance as yf
ticker_obj = yf.Ticker(ticker)
info = ticker_obj.info
financials = ticker_obj.financials
balance_sheet = ticker_obj.balance_sheet
cashflow = ticker_obj.cashflow
# Compose a context string
context = f"""
# Company Info
Name: {info.get('shortName', 'N/A')}
Industry: {info.get('industry', 'N/A')}
Sector: {info.get('sector', 'N/A')}
Country: {info.get('country', 'N/A')}
Website: {info.get('website', 'N/A')}
response = client.responses.create( # Financials (last available)
model=config["quick_think_llm"], {financials.to_string() if not financials.empty else 'N/A'}
input=[
{
"role": "system",
"content": [
{
"type": "input_text",
"text": f"Can you search Fundamental for discussions on {ticker} during of the month before {curr_date} to the month of {curr_date}. Make sure you only get the data posted during that period. List as a table, with PE/PS/Cash flow/ etc",
}
],
}
],
text={"format": {"type": "text"}},
reasoning={},
tools=[
{
"type": "web_search_preview",
"user_location": {"type": "approximate"},
"search_context_size": "low",
}
],
temperature=1,
max_output_tokens=4096,
top_p=1,
store=True,
)
return response.output[1].content[0].text # Balance Sheet (last available)
{balance_sheet.to_string() if not balance_sheet.empty else 'N/A'}
# Cash Flow (last available)
{cashflow.to_string() if not cashflow.empty else 'N/A'}
"""
prompt = (
f"Given the following real company data for {ticker} as of {curr_date}, "
"write a detailed fundamental analysis. Include company profile, financials, PE/PS/Cash flow, and any recent news or events that would affect fundamentals. "
"Present the information in a markdown table at the end.\n\n"
f"{context}"
)
if provider == "google":
google_api_key = os.getenv("GOOGLE_API_KEY")
llm = ChatGoogleGenerativeAI(model=config["quick_think_llm"], google_api_key=google_api_key)
result = llm.invoke(prompt)
return result.content
elif provider == "anthropic":
anthropic_api_key = os.getenv("ANTHROPIC_API_KEY")
llm = ChatAnthropic(model_name=config["quick_think_llm"], api_key=anthropic_api_key)
result = llm.invoke(prompt)
return result.content
else:
return "Fundamental data retrieval is not supported for this LLM provider."

5
tradingagents/graph/trading_graph.py
View File

@ -65,8 +65,9 @@ class TradingAgentsGraph:
self.deep_thinking_llm = ChatAnthropic(model_name=self.config["deep_think_llm"], base_url=self.config["backend_url"], timeout=120, stop=None) self.deep_thinking_llm = ChatAnthropic(model_name=self.config["deep_think_llm"], base_url=self.config["backend_url"], timeout=120, stop=None)
self.quick_thinking_llm = ChatAnthropic(model_name=self.config["quick_think_llm"], base_url=self.config["backend_url"], timeout=120, stop=None) self.quick_thinking_llm = ChatAnthropic(model_name=self.config["quick_think_llm"], base_url=self.config["backend_url"], timeout=120, stop=None)
elif self.config["llm_provider"].lower() == "google": elif self.config["llm_provider"].lower() == "google":
self.deep_thinking_llm = ChatGoogleGenerativeAI(model=self.config["deep_think_llm"]) google_api_key = os.getenv("GOOGLE_API_KEY")
self.quick_thinking_llm = ChatGoogleGenerativeAI(model=self.config["quick_think_llm"]) self.deep_thinking_llm = ChatGoogleGenerativeAI(model=self.config["deep_think_llm"], google_api_key=google_api_key)
self.quick_thinking_llm = ChatGoogleGenerativeAI(model=self.config["quick_think_llm"], google_api_key=google_api_key)
else: else:
raise ValueError(f"Unsupported LLM provider: {self.config['llm_provider']}") raise ValueError(f"Unsupported LLM provider: {self.config['llm_provider']}")