PrimoGreedy-Agent / src /tools /daily_csv_tool.py
CiscsoPonce's picture
Initial Deploy (Clean)
a2cbcac
import os
import pandas as pd
from datetime import datetime
from typing import Dict, Any, List, Union
from pathlib import Path
from ..config import config
def format_number_to_2_decimals(value: Union[int, float, str, None]) -> Union[str, None]:
"""
Format numeric values to 2 decimal places.
Args:
value: The value to format
Returns:
Formatted string with 2 decimals, or original value if not numeric
"""
if value is None:
return None
# Keep string values like 'N/A' unchanged
if isinstance(value, str):
return value
# Format numeric values to 2 decimal places
if isinstance(value, (int, float)):
return f"{value:.2f}"
return value
def save_workflow_to_csv(workflow_result: Dict[str, Any], date: str) -> bool:
"""
Save workflow results to a single CSV file with each row representing a different date.
Args:
workflow_result: Complete workflow result containing all analysis data
date: Date string for the analysis
Returns:
bool: True if saved successfully, False otherwise
"""
try:
if not workflow_result.get('success', False):
print(f"No data to save to CSV.")
return False
# Extract results from workflow
results = workflow_result.get('results', {})
symbols = workflow_result.get('symbols', [])
if not symbols:
print(f"No symbols found in workflow result.")
return False
symbol = symbols[0] # We process one symbol at a time
# Get portfolio manager results - Access through symbol key
portfolio_results = results.get('portfolio_manager', {})
if not portfolio_results or symbol not in portfolio_results:
print(f"No portfolio manager results found for {symbol}.")
return False
symbol_portfolio_result = portfolio_results[symbol]
# Check if portfolio analysis was successful
if not symbol_portfolio_result.get('success', False):
print(f"Portfolio analysis failed for {symbol}: {symbol_portfolio_result.get('error', 'Unknown error')}")
return False
# Get data collection results for current price
data_results = results.get('data_collection', {})
market_data = data_results.get('market_data', {}) if data_results else {}
current_price = market_data.get('current_price', 'N/A')
# Get technical analysis results
tech_results = results.get('technical_analysis', {})
tech_indicators = {}
if tech_results and tech_results.get('success'):
indicators_data = tech_results.get('indicators', {})
tech_indicators = indicators_data.get('technical_indicators', {})
# Helper function to get last value from indicator lists
def get_last_value(data):
if isinstance(data, list) and data:
return data[-1]
elif isinstance(data, (int, float)):
return data
return None
# Extract individual technical indicators
sma = get_last_value(tech_indicators.get('SMA', []))
rsi = get_last_value(tech_indicators.get('RSI', []))
adx = get_last_value(tech_indicators.get('ADX', []))
cci = get_last_value(tech_indicators.get('CCI', []))
# MACD components
macd_data = tech_indicators.get('MACD', {})
macd_line = get_last_value(macd_data.get('macd', [])) if isinstance(macd_data, dict) else None
macd_signal = get_last_value(macd_data.get('signal', [])) if isinstance(macd_data, dict) else None
macd_histogram = get_last_value(macd_data.get('histogram', [])) if isinstance(macd_data, dict) else None
# Bollinger Bands components
bb_data = tech_indicators.get('BBANDS', {})
bb_upper = get_last_value(bb_data.get('upper', [])) if isinstance(bb_data, dict) else None
bb_middle = get_last_value(bb_data.get('middle', [])) if isinstance(bb_data, dict) else None
bb_lower = get_last_value(bb_data.get('lower', [])) if isinstance(bb_data, dict) else None
# Get news intelligence results for NLP features
news_results = results.get('news_intelligence', {})
nlp_features = {}
if news_results and news_results.get('success'):
nlp_features = news_results.get('nlp_features', {})
# Create comprehensive CSV data with all indicators and NLP features
# Apply 2-decimal formatting ONLY to technical/financial fields
csv_data = {
# Basic data
'date': date,
'symbol': symbol,
'close': format_number_to_2_decimals(current_price),
'trading_signal': symbol_portfolio_result.get('trading_signal', 'N/A'),
'confidence_level': symbol_portfolio_result.get('confidence_level', 'N/A'), # Keep original precision
'position_size': symbol_portfolio_result.get('position_size', 'N/A'), # Keep original precision
# Technical indicators - format to 2 decimals
'sma': format_number_to_2_decimals(sma),
'rsi': format_number_to_2_decimals(rsi),
'adx': format_number_to_2_decimals(adx),
'cci': format_number_to_2_decimals(cci),
# MACD components - format to 2 decimals
'macd_line': format_number_to_2_decimals(macd_line),
'macd_signal': format_number_to_2_decimals(macd_signal),
'macd_histogram': format_number_to_2_decimals(macd_histogram),
# Bollinger Bands components - format to 2 decimals
'bb_upper': format_number_to_2_decimals(bb_upper),
'bb_middle': format_number_to_2_decimals(bb_middle),
'bb_lower': format_number_to_2_decimals(bb_lower),
# NLP features from news intelligence - keep original precision
'news_relevance': nlp_features.get('news_relevance'),
'sentiment': nlp_features.get('sentiment'),
'price_impact_potential': nlp_features.get('price_impact_potential'),
'trend_direction': nlp_features.get('trend_direction'),
'earnings_impact': nlp_features.get('earnings_impact'),
'investor_confidence': nlp_features.get('investor_confidence'),
'risk_profile_change': nlp_features.get('risk_profile_change')
}
# Create CSV file path - SINGLE FILE for all dates
csv_path = config.csv_output_path
os.makedirs(csv_path, exist_ok=True)
csv_file = os.path.join(csv_path, 'daily_analysis.csv')
# Check if file exists and load existing data
file_exists = os.path.exists(csv_file)
if file_exists:
# Load existing data and append new row
try:
existing_df = pd.read_csv(csv_file)
new_df = pd.DataFrame([csv_data])
combined_df = pd.concat([existing_df, new_df], ignore_index=True)
# Remove duplicate entries for same date/symbol, keeping the latest
combined_df = combined_df.drop_duplicates(subset=['date', 'symbol'], keep='last')
# Sort by date descending (newest first)
combined_df = combined_df.sort_values(by='date', ascending=False)
except pd.errors.EmptyDataError:
# File exists but is empty
combined_df = pd.DataFrame([csv_data])
else:
# Create new file
combined_df = pd.DataFrame([csv_data])
# Save to CSV
combined_df.to_csv(csv_file, index=False)
print(f"CSV saved successfully: {csv_file}")
print(f"Total rows in CSV: {len(combined_df)}")
return True
except Exception as e:
print(f"Error saving CSV: {e}")
import traceback
traceback.print_exc()
return False
def save_workflow_to_symbol_csv(workflow_result: Dict[str, Any], date: str, data_dir: str = "./output/csv") -> bool:
"""
Save workflow results to a per-symbol CSV under data_dir with filename
pattern: daily_analysis_{SYMBOL}.csv.
The file will include at least: date, symbol, close, trading_signal,
confidence_level, position_size, and selected indicators and NLP features.
Args:
workflow_result: Workflow result dict
date: Analysis date (YYYY-MM-DD)
data_dir: Target directory for symbol CSV files (default ./data)
Returns:
bool: True if saved successfully
"""
try:
if not workflow_result.get('success', False):
return False
results = workflow_result.get('results', {})
symbols = workflow_result.get('symbols', [])
if not symbols:
return False
symbol = symbols[0].upper()
portfolio_results = results.get('portfolio_manager', {})
if not portfolio_results or symbol not in portfolio_results:
return False
symbol_portfolio_result = portfolio_results[symbol]
if not symbol_portfolio_result.get('success', False):
return False
data_results = results.get('data_collection', {})
market_data = data_results.get('market_data', {}) if data_results else {}
current_price = market_data.get('current_price', 'N/A')
tech_results = results.get('technical_analysis', {})
tech_indicators = {}
if tech_results and tech_results.get('success'):
indicators_data = tech_results.get('indicators', {})
tech_indicators = indicators_data.get('technical_indicators', {})
def get_last_value(data):
if isinstance(data, list) and data:
return data[-1]
elif isinstance(data, (int, float)):
return data
return None
sma = get_last_value(tech_indicators.get('SMA', []))
rsi = get_last_value(tech_indicators.get('RSI', []))
adx = get_last_value(tech_indicators.get('ADX', []))
cci = get_last_value(tech_indicators.get('CCI', []))
macd_data = tech_indicators.get('MACD', {})
macd_line = get_last_value(macd_data.get('macd', [])) if isinstance(macd_data, dict) else None
macd_signal = get_last_value(macd_data.get('signal', [])) if isinstance(macd_data, dict) else None
macd_histogram = get_last_value(macd_data.get('histogram', [])) if isinstance(macd_data, dict) else None
bb_data = tech_indicators.get('BBANDS', {})
bb_upper = get_last_value(bb_data.get('upper', [])) if isinstance(bb_data, dict) else None
bb_middle = get_last_value(bb_data.get('middle', [])) if isinstance(bb_data, dict) else None
bb_lower = get_last_value(bb_data.get('lower', [])) if isinstance(bb_data, dict) else None
news_results = results.get('news_intelligence', {})
nlp_features = news_results.get('nlp_features', {}) if news_results and news_results.get('success') else {}
csv_row = {
'date': date,
'symbol': symbol,
'close': format_number_to_2_decimals(current_price),
'trading_signal': symbol_portfolio_result.get('trading_signal', 'N/A'),
'confidence_level': symbol_portfolio_result.get('confidence_level', 'N/A'),
'position_size': symbol_portfolio_result.get('position_size', 'N/A'),
'sma': format_number_to_2_decimals(sma),
'rsi': format_number_to_2_decimals(rsi),
'adx': format_number_to_2_decimals(adx),
'cci': format_number_to_2_decimals(cci),
'macd_line': format_number_to_2_decimals(macd_line),
'macd_signal': format_number_to_2_decimals(macd_signal),
'macd_histogram': format_number_to_2_decimals(macd_histogram),
'bb_upper': format_number_to_2_decimals(bb_upper),
'bb_middle': format_number_to_2_decimals(bb_middle),
'bb_lower': format_number_to_2_decimals(bb_lower),
'news_relevance': nlp_features.get('news_relevance'),
'sentiment': nlp_features.get('sentiment'),
'price_impact_potential': nlp_features.get('price_impact_potential'),
'trend_direction': nlp_features.get('trend_direction'),
'earnings_impact': nlp_features.get('earnings_impact'),
'investor_confidence': nlp_features.get('investor_confidence'),
'risk_profile_change': nlp_features.get('risk_profile_change'),
}
# Ensure directory and file path
Path(data_dir).mkdir(parents=True, exist_ok=True)
symbol_csv = Path(data_dir) / f"daily_analysis_{symbol}.csv"
# Append/update row
if symbol_csv.exists():
try:
existing_df = pd.read_csv(symbol_csv)
new_df = pd.DataFrame([csv_row])
combined = pd.concat([existing_df, new_df], ignore_index=True)
combined = combined.drop_duplicates(subset=['date'], keep='last')
combined = combined.sort_values(by='date', ascending=True)
except pd.errors.EmptyDataError:
combined = pd.DataFrame([csv_row])
else:
combined = pd.DataFrame([csv_row])
combined.to_csv(symbol_csv, index=False)
print(f"Symbol CSV saved: {symbol_csv}")
return True
except Exception as e:
print(f"Error saving symbol CSV: {e}")
return False