Spaces:
Running
Running
File size: 30,224 Bytes
26c1feb 80bf6fa cb76f6b 0df06c7 cb76f6b 0df06c7 cb76f6b a6be62a 80bf6fa |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 |
import requests
from bs4 import BeautifulSoup
from transformers import AutoTokenizer, AutoModelForSequenceClassification
from transformers import pipeline
from collections import Counter
import time
import numpy as np
import yfinance as yf
import pandas as pd
from datetime import datetime, timedelta
import json
from typing import Dict, List, Tuple
import re # Add this import
import warnings
warnings.filterwarnings('ignore')
# Load FinBERT
model_name = "yiyanghkust/finbert-tone"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
classifier = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)
class StockSentimentAnalyzer:
def __init__(self):
self.session = requests.Session()
self.session.headers.update({
'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36'
})
# API setup for Indian stock data
self.api_url = "https://indian-stock-exchange-api2.p.rapidapi.com/stock"
self.api_headers = {
"x-rapidapi-host": "indian-stock-exchange-api2.p.rapidapi.com",
"x-rapidapi-key": "a12f59fc40msh153da8fdf3885b6p100406jsn57d1d84b0d06"
}
self.symbol = None
def get_stock_data(self, symbol: str, period: str = "1mo") -> pd.DataFrame:
"""Fetch stock data from Yahoo Finance"""
try:
# Add .NS for NSE stocks if not present
if not symbol.endswith('.NS') and not symbol.endswith('.BO'):
symbol += '.NS'
stock = yf.Ticker(symbol)
data = stock.history(period=period)
return data
except Exception as e:
print(f"Error fetching stock data for {symbol}: {e}")
return pd.DataFrame()
def get_news_from_api(self, company_name: str) -> List[Dict]:
"""Get news articles from the API"""
querystring = {"name": company_name}
try:
response = requests.get(self.api_url, headers=self.api_headers, params=querystring)
data = response.json()
news_data = data.get("recentNews", {})
return news_data
except Exception as e:
print(f"Error fetching news from API: {e}")
return []
def scrape_news_sentiment(self, company_name: str, symbol: str) -> Dict:
"""Scrape news sentiment from multiple sources"""
news_data = {
'headlines': [],
'sources': [],
'sentiment_scores': [],
'dates': [],
'urls': []
}
# Get news from API
api_news = self.get_news_from_api(company_name)
urls = [item["url"] for item in api_news if isinstance(item, dict) and "url" in item]
print(f"Found {len(urls)} news articles from API")
# Process each URL
for i, news_url in enumerate(urls):
try:
print(f"\n[{i+1}/{len(urls)}] Analyzing: {news_url[:60]}...")
html = requests.get(news_url, timeout=10).text
soup = BeautifulSoup(html, "html.parser")
# Get title
title = soup.title.string if soup.title else "No title"
# Grab <p> tags and filter
paragraphs = soup.find_all("p")
if not paragraphs:
print("โ No content found")
continue
content = " ".join(p.get_text() for p in paragraphs if len(p.get_text()) > 40)
content = content.strip()
if len(content) < 100:
print("โ Content too short")
continue
# Truncate to 512 tokens max
content = content[:1000]
result = classifier(content[:512])[0]
label = result['label'].lower()
score = result['score']
# Convert FinBERT sentiment to polarity score (-1 to 1)
polarity = 0
if label == "positive":
polarity = score
elif label == "negative":
polarity = -score
news_data['headlines'].append(title)
news_data['sources'].append('API')
news_data['sentiment_scores'].append(polarity)
news_data['dates'].append(datetime.now())
news_data['urls'].append(news_url)
print(f"โ Sentiment: {label.upper()} (confidence: {score:.1%})")
time.sleep(1.2) # polite delay
except Exception as e:
print(f"โ Error: {str(e)}")
continue
# Economic Times
try:
et_url = f"https://economictimes.indiatimes.com/topic/{company_name.replace(' ', '-')}"
response = self.session.get(et_url, timeout=10)
soup = BeautifulSoup(response.content, 'html.parser')
headlines = soup.find_all(['h2', 'h3', 'h4'], class_=re.compile('.*title.*|.*headline.*'))
for headline in headlines[:5]: # Limit to 5 headlines
text = headline.get_text().strip()
if text and len(text) > 10:
# Use FinBERT for sentiment analysis
result = classifier(text)[0]
label = result['label'].lower()
score = result['score']
# Convert to polarity
polarity = 0
if label == "positive":
polarity = score
elif label == "negative":
polarity = -score
news_data['headlines'].append(text)
news_data['sources'].append('Economic Times')
news_data['sentiment_scores'].append(polarity)
news_data['dates'].append(datetime.now())
news_data['urls'].append(et_url)
except Exception as e:
print(f"Error scraping Economic Times: {e}")
return news_data
def calculate_news_sentiment_score(self, news_data: Dict) -> Dict:
"""Calculate various sentiment scores from news data"""
if not news_data['sentiment_scores']:
return {
'positive_score': 50,
'negative_score': 50,
'fear_score': 50,
'confidence_score': 50,
'overall_sentiment_score': 50
}
sentiments = news_data['sentiment_scores']
headlines = news_data['headlines']
# Count sentiments
positive_count = sum(1 for s in sentiments if s > 0.1)
negative_count = sum(1 for s in sentiments if s < -0.1)
neutral_count = len(sentiments) - positive_count - negative_count
total = len(sentiments)
positive_score = (positive_count / total) * 100 if total > 0 else 50
negative_score = (negative_count / total) * 100 if total > 0 else 50
# Calculate average confidence
confidence_values = [abs(s) for s in sentiments]
avg_confidence = sum(confidence_values) / len(confidence_values) if confidence_values else 0
confidence_score = avg_confidence * 100
# Fear score based on keywords
fear_keywords = ['fall', 'drop', 'crash', 'loss', 'decline', 'bear', 'sell', 'down', 'negative', 'risk']
confidence_keywords = ['rise', 'gain', 'bull', 'buy', 'up', 'positive', 'growth', 'profit', 'strong']
fear_mentions = sum(1 for headline in headlines
for keyword in fear_keywords
if keyword.lower() in headline.lower())
confidence_mentions = sum(1 for headline in headlines
for keyword in confidence_keywords
if keyword.lower() in headline.lower())
fear_score = min(100, (fear_mentions / len(headlines)) * 200) if headlines else 50
confidence_boost = min(100, (confidence_mentions / len(headlines)) * 200) if headlines else 50
# Overall sentiment score
overall_sentiment = 50 + ((positive_score - negative_score) * 0.3) + ((confidence_boost - fear_score) * 0.2)
return {
'positive_score': round(positive_score, 2),
'negative_score': round(negative_score, 2),
'fear_score': round(fear_score, 2),
'confidence_score': round(confidence_score, 2),
'overall_sentiment_score': round(min(100, max(0, overall_sentiment)), 2)
}
def calculate_volatility_score(self, stock_data: pd.DataFrame) -> float:
"""Calculate innovative volatility score (0-100)"""
if stock_data.empty:
return 0
# Calculate different volatility measures
returns = stock_data['Close'].pct_change().dropna()
# Standard deviation of returns (annualized)
std_vol = returns.std() * np.sqrt(252) * 100
# Average True Range volatility
high_low = stock_data['High'] - stock_data['Low']
high_close = np.abs(stock_data['High'] - stock_data['Close'].shift())
low_close = np.abs(stock_data['Low'] - stock_data['Close'].shift())
true_range = np.maximum(high_low, np.maximum(high_close, low_close))
atr = true_range.rolling(14).mean().iloc[-1]
atr_vol = (atr / stock_data['Close'].iloc[-1]) * 100
# Price range volatility
price_range = ((stock_data['High'].max() - stock_data['Low'].min()) / stock_data['Close'].iloc[-1]) * 100
# Combine and normalize to 0-100 scale
volatility_score = min(100, (std_vol * 0.4 + atr_vol * 0.4 + price_range * 0.2))
return round(volatility_score, 2)
def calculate_momentum_score(self, stock_data: pd.DataFrame) -> float:
"""Calculate momentum score based on price trends (0-100)"""
if stock_data.empty:
return 50
# RSI calculation
delta = stock_data['Close'].diff()
gain = (delta.where(delta > 0, 0)).rolling(window=14).mean()
loss = (-delta.where(delta < 0, 0)).rolling(window=14).mean()
rs = gain / loss
rsi = 100 - (100 / (1 + rs))
current_rsi = rsi.iloc[-1] if not np.isnan(rsi.iloc[-1]) else 50
# Price momentum (% change over different periods)
mom_1d = ((stock_data['Close'].iloc[-1] - stock_data['Close'].iloc[-2]) / stock_data['Close'].iloc[-2]) * 100
mom_5d = ((stock_data['Close'].iloc[-1] - stock_data['Close'].iloc[-6]) / stock_data['Close'].iloc[-6]) * 100 if len(stock_data) > 5 else 0
mom_20d = ((stock_data['Close'].iloc[-1] - stock_data['Close'].iloc[-21]) / stock_data['Close'].iloc[-21]) * 100 if len(stock_data) > 20 else 0
# Moving average trends
ma_5 = stock_data['Close'].rolling(5).mean().iloc[-1]
ma_20 = stock_data['Close'].rolling(20).mean().iloc[-1] if len(stock_data) > 20 else ma_5
current_price = stock_data['Close'].iloc[-1]
ma_score = 50
if current_price > ma_5 > ma_20:
ma_score = 75
elif current_price > ma_5:
ma_score = 65
elif current_price < ma_5 < ma_20:
ma_score = 25
elif current_price < ma_5:
ma_score = 35
# Combine scores
momentum_score = (current_rsi * 0.4 + ma_score * 0.3 +
min(max(mom_1d * 2 + 50, 0), 100) * 0.1 +
min(max(mom_5d + 50, 0), 100) * 0.1 +
min(max(mom_20d * 0.5 + 50, 0), 100) * 0.1)
return round(momentum_score, 2)
def calculate_liquidity_score(self, stock_data: pd.DataFrame) -> float:
"""Calculate liquidity score based on volume patterns (0-100)"""
if stock_data.empty:
return 0
# Average volume
avg_volume = stock_data['Volume'].mean()
recent_volume = stock_data['Volume'].tail(5).mean()
# Volume trend
volume_trend = (recent_volume - avg_volume) / avg_volume * 100 if avg_volume > 0 else 0
# Volume-price relationship
price_changes = stock_data['Close'].pct_change()
volume_changes = stock_data['Volume'].pct_change()
correlation = price_changes.corr(volume_changes)
correlation = 0 if np.isnan(correlation) else correlation
# Normalize to 0-100 scale
volume_score = min(100, max(0, 50 + volume_trend * 0.3 + correlation * 25))
return round(volume_score, 2)
def calculate_technical_strength_score(self, stock_data: pd.DataFrame) -> float:
"""Calculate technical strength based on multiple indicators (0-100)"""
if stock_data.empty:
return 50
scores = []
# Support/Resistance levels
highs = stock_data['High'].rolling(20).max()
lows = stock_data['Low'].rolling(20).min()
current_price = stock_data['Close'].iloc[-1]
# Price position within range
price_position = ((current_price - lows.iloc[-1]) / (highs.iloc[-1] - lows.iloc[-1])) * 100
scores.append(min(100, max(0, price_position)))
# Volume-weighted average price deviation
vwap = (stock_data['Close'] * stock_data['Volume']).sum() / stock_data['Volume'].sum()
vwap_score = 50 + ((current_price - vwap) / vwap) * 100
scores.append(min(100, max(0, vwap_score)))
# Bollinger Bands position
ma_20 = stock_data['Close'].rolling(20).mean()
std_20 = stock_data['Close'].rolling(20).std()
upper_band = ma_20 + (std_20 * 2)
lower_band = ma_20 - (std_20 * 2)
if not upper_band.empty and not lower_band.empty:
bb_position = ((current_price - lower_band.iloc[-1]) /
(upper_band.iloc[-1] - lower_band.iloc[-1])) * 100
scores.append(min(100, max(0, bb_position)))
return round(np.mean(scores), 2)
def calculate_market_correlation_score(self, symbol: str, stock_data: pd.DataFrame) -> float:
"""Calculate correlation with major indices (0-100)"""
try:
# Get Nifty 50 data for comparison
nifty = yf.Ticker("^NSEI")
nifty_data = nifty.history(period="1mo")
if nifty_data.empty or stock_data.empty:
return 50
# Align dates
common_dates = stock_data.index.intersection(nifty_data.index)
if len(common_dates) < 5:
return 50
stock_returns = stock_data.loc[common_dates]['Close'].pct_change().dropna()
nifty_returns = nifty_data.loc[common_dates]['Close'].pct_change().dropna()
# Calculate correlation
correlation = stock_returns.corr(nifty_returns)
if np.isnan(correlation):
return 50
# Convert correlation to 0-100 score
# High positive correlation = higher score (follows market)
# Negative correlation = lower score (contrarian)
correlation_score = (correlation + 1) * 50
return round(correlation_score, 2)
except Exception as e:
print(f"Error calculating market correlation: {e}")
return 50
def calculate_growth_potential_score(self, stock_data: pd.DataFrame) -> float:
"""Calculate growth potential based on trend analysis (0-100)"""
if stock_data.empty:
return 50
# Calculate different timeframe growth rates
current_price = stock_data['Close'].iloc[-1]
growth_scores = []
# Weekly growth
if len(stock_data) >= 5:
week_ago_price = stock_data['Close'].iloc[-5]
weekly_growth = ((current_price - week_ago_price) / week_ago_price) * 100
weekly_score = min(100, max(0, 50 + weekly_growth * 2))
growth_scores.append(weekly_score)
# Monthly growth
if len(stock_data) >= 20:
month_ago_price = stock_data['Close'].iloc[-20]
monthly_growth = ((current_price - month_ago_price) / month_ago_price) * 100
monthly_score = min(100, max(0, 50 + monthly_growth))
growth_scores.append(monthly_score)
# Volume growth trend
recent_volume = stock_data['Volume'].tail(5).mean()
earlier_volume = stock_data['Volume'].head(5).mean()
if earlier_volume > 0:
volume_growth = ((recent_volume - earlier_volume) / earlier_volume) * 100
volume_score = min(100, max(0, 50 + volume_growth * 0.5))
growth_scores.append(volume_score)
return round(np.mean(growth_scores) if growth_scores else 50, 2)
def calculate_stability_score(self, stock_data: pd.DataFrame) -> float:
"""Calculate stability score based on price steadiness (0-100)"""
if stock_data.empty:
return 50
# Calculate coefficient of variation
returns = stock_data['Close'].pct_change().dropna()
mean_return = returns.mean()
std_return = returns.std()
if mean_return != 0:
cv = abs(std_return / mean_return)
# Lower CV = higher stability
stability_score = max(0, 100 - cv * 100)
else:
stability_score = 50
# Consider price gaps
gaps = abs(stock_data['Open'] - stock_data['Close'].shift()).dropna()
avg_gap = gaps.mean()
avg_price = stock_data['Close'].mean()
if avg_price > 0:
gap_ratio = avg_gap / avg_price
gap_penalty = min(50, gap_ratio * 1000)
stability_score = max(0, stability_score - gap_penalty)
return round(stability_score, 2)
def calculate_risk_score(self, analysis: Dict) -> float:
"""Calculate risk score based on multiple factors"""
risk_factors = [
analysis['volatility_score'],
analysis['fear_score'],
100 - analysis['liquidity_score'],
100 - analysis['technical_strength_score'],
100 - analysis['stability_score']
]
return round(np.mean(risk_factors), 2)
def calculate_investment_attractiveness(self, analysis: Dict) -> float:
"""Calculate investment attractiveness score"""
attractiveness_factors = [
analysis['overall_sentiment_score'],
analysis['growth_potential_score'],
analysis['momentum_score'],
100 - analysis['risk_score']
]
return round(np.mean(attractiveness_factors), 2)
def get_comprehensive_analysis(self, symbol: str, company_name: str = None) -> Dict:
"""Get comprehensive sentiment analysis for a stock"""
# If company name not provided, try to extract from symbol
self.symbol = symbol
# If company name not provided, try to extract from symbol
if not company_name:
try:
# Add .NS for NSE stocks if not present
if not symbol.endswith('.NS') and not symbol.endswith('.BO'):
symbol_with_suffix = symbol + '.NS'
else:
symbol_with_suffix = symbol
# Get company info from yfinance
ticker = yf.Ticker(symbol_with_suffix)
info = ticker.info
# Extract company name with multiple fallbacks
company_name = (
info.get('longName') or
info.get('shortName') or
info.get('name') or
symbol # Final fallback to symbol
)
# Validate the extracted name
if company_name:
# Remove special characters and check if meaningful
cleaned_name = ''.join(c for c in company_name if c.isalnum() or c in (' ', '-', '&'))
if (len(cleaned_name.strip()) < 2 or # Too short
cleaned_name.strip() == symbol or # Same as symbol
any(x in cleaned_name for x in ['-', ' - ']) or # Contains dashes (likely placeholder)
cleaned_name.isnumeric()): # Just numbers
company_name = symbol # Fallback to symbol if name is invalid
else:
company_name = symbol
except Exception as e:
print(f"โ ๏ธ Could not fetch company name for {symbol}: {str(e)}")
company_name = symbol
# Ensure we have at least the symbol as name
company_name = company_name or symbol
print(company_name)
print(f"\n{'='*80}")
print(f"๐ ANALYZING: {company_name.upper()} ({symbol})")
print(f"{'='*80}")
# Get stock data
print("๐ Fetching stock data...")
stock_data = self.get_stock_data(symbol)
if stock_data.empty:
print("โ Could not fetch stock data. Please check the symbol.")
return {}
# Get news sentiment
print("๐ฐ Scraping news sentiment...")
news_data = self.scrape_news_sentiment(company_name, symbol)
# Calculate all scores
print("๐งฎ Calculating sentiment scores...")
# Basic stock info
current_price = stock_data['Close'].iloc[-1]
prev_close = stock_data['Close'].iloc[-2] if len(stock_data) > 1 else current_price
price_change = current_price - prev_close
price_change_pct = (price_change / prev_close) * 100 if prev_close != 0 else 0
analysis = {
'symbol': symbol,
'company_name': company_name,
'analysis_date': datetime.now().strftime('%Y-%m-%d %H:%M:%S'),
'current_price': round(current_price, 2),
'price_change': round(price_change, 2),
'price_change_pct': round(price_change_pct, 2),
'volume': int(stock_data['Volume'].iloc[-1]),
'market_cap_approx': 'N/A', # Would need additional API for exact market cap
# Innovative Scores
'volatility_score': self.calculate_volatility_score(stock_data),
'momentum_score': self.calculate_momentum_score(stock_data),
'liquidity_score': self.calculate_liquidity_score(stock_data),
'technical_strength_score': self.calculate_technical_strength_score(stock_data),
'market_correlation_score': self.calculate_market_correlation_score(symbol, stock_data),
'growth_potential_score': self.calculate_growth_potential_score(stock_data),
'stability_score': self.calculate_stability_score(stock_data),
# News sentiment scores
**self.calculate_news_sentiment_score(news_data),
# Additional metrics
'news_count': len(news_data['headlines']),
'recent_headlines': news_data['headlines'][:5] # Top 5 headlines
}
# Calculate risk score
analysis['risk_score'] = self.calculate_risk_score(analysis)
# Calculate risk level based on risk score
if analysis['risk_score'] >= 75:
analysis['risk_level'] = "VERY HIGH"
elif analysis['risk_score'] >= 60:
analysis['risk_level'] = "HIGH"
elif analysis['risk_score'] >= 40:
analysis['risk_level'] = "MODERATE"
elif analysis['risk_score'] >= 25:
analysis['risk_level'] = "LOW"
else:
analysis['risk_level'] = "VERY LOW"
# Add risk factors based on analysis
analysis['risk_factors'] = []
if analysis['volatility_score'] > 70:
analysis['risk_factors'].append("High market volatility")
if analysis['fear_score'] > 60:
analysis['risk_factors'].append("Elevated market fear")
if analysis['negative_score'] > 60:
analysis['risk_factors'].append("Negative sentiment trend")
if analysis['market_correlation_score'] < 30:
analysis['risk_factors'].append("Low market correlation")
if analysis['stability_score'] < 40:
analysis['risk_factors'].append("Low stability indicators")
# Calculate investment attractiveness
analysis['investment_attractiveness_score'] = self.calculate_investment_attractiveness(analysis)
return analysis
def generate_recommendation(self, analysis: Dict) -> str:
"""Generate trading recommendation based on analysis"""
if not analysis:
return "Unable to generate recommendation - insufficient data"
sentiment = analysis['overall_sentiment_score']
risk = analysis['risk_score']
momentum = analysis['momentum_score']
volatility = analysis['volatility_score']
attractiveness = analysis['investment_attractiveness_score']
if sentiment > 70 and risk < 40 and momentum > 60 and attractiveness > 65:
return "๐ข STRONG BUY - High sentiment, low risk, strong momentum"
elif sentiment > 60 and risk < 50 and attractiveness > 55:
return "๐ข BUY - Positive sentiment with manageable risk"
elif sentiment > 40 and sentiment < 60 and risk < 60:
return "๐ก HOLD - Neutral sentiment, monitor closely"
elif sentiment < 40 and risk > 60:
return "๐ด SELL - Negative sentiment with high risk"
elif sentiment < 30 or risk > 75:
return "๐ด STRONG SELL - Very negative sentiment or very high risk"
else:
return "๐ก HOLD - Mixed signals, proceed with caution"
def display_analysis(self, analysis: Dict):
"""Display comprehensive analysis in a formatted way"""
if not analysis:
print("โ No analysis data available")
return
print(f"\n{'='*80}")
print(f"๐ COMPREHENSIVE STOCK ANALYSIS REPORT")
print(f"{'='*80}")
# Basic Info
print(f"\n๐ BASIC INFORMATION:")
print(f"Company: {analysis['company_name']}")
print(f"Symbol: {analysis['symbol']}")
print(f"Current Price: โน{analysis['current_price']}")
print(f"Price Change: โน{analysis['price_change']} ({analysis['price_change_pct']:+.2f}%)")
print(f"Volume: {analysis['volume']:,}")
print(f"Analysis Date: {analysis['analysis_date']}")
# Sentiment Scores
print(f"\n๐ฏ SENTIMENT SCORES (0-100):")
print(f"Overall Sentiment Score: {analysis['overall_sentiment_score']}/100")
print(f"Positive Score: {analysis['positive_score']}/100")
print(f"Negative Score: {analysis['negative_score']}/100")
print(f"Fear Score: {analysis['fear_score']}/100")
print(f"Confidence Score: {analysis['confidence_score']}/100")
# Technical Scores
print(f"\nโ๏ธ TECHNICAL SCORES (0-100):")
print(f"Volatility Score: {analysis['volatility_score']}/100")
print(f"Momentum Score: {analysis['momentum_score']}/100")
print(f"Technical Strength: {analysis['technical_strength_score']}/100")
print(f"Liquidity Score: {analysis['liquidity_score']}/100")
print(f"Market Correlation: {analysis['market_correlation_score']}/100")
# Advanced Scores
print(f"\n๐ ADVANCED SCORES (0-100):")
print(f"Growth Potential: {analysis['growth_potential_score']}/100")
print(f"Stability Score: {analysis['stability_score']}/100")
print(f"Risk Score: {analysis['risk_score']}/100")
print(f"Investment Attractiveness: {analysis['investment_attractiveness_score']}/100")
# Recommendation
recommendation = self.generate_recommendation(analysis)
print(f"\n๐ก RECOMMENDATION:")
print(f"{recommendation}")
# News Analysis
print(f"\n๐ฐ NEWS ANALYSIS:")
print(f"Headlines Analyzed: {analysis['news_count']}")
if analysis['recent_headlines']:
print(f"\n๐ Recent Headlines:")
for i, headline in enumerate(analysis['recent_headlines'], 1):
print(f"{i}. {headline}")
# Risk Assessment
print(f"\nโ ๏ธ RISK ASSESSMENT:")
print(f"Risk Level: {analysis['risk_level']}")
print(f"Key Risk Factors:")
for risk_factor in analysis['risk_factors']:
print(f"- {risk_factor}")
# Save analysis to JSON
output_file = f"analysis_{self.symbol}_{datetime.now().strftime('%Y%m%d_%H%M%S')}.json"
with open(output_file, 'w') as f:
json.dump(analysis, f, indent=4)
print(f"\n๐พ Analysis saved to {output_file}")
def main():
"""Main function to run the stock analysis"""
analyzer = StockSentimentAnalyzer()
print("๐ Welcome to Stock Sentiment Analyzer!")
print("Enter stock symbols (e.g., RELIANCE, TCS, HDFCBANK)")
print("The system will automatically add .NS for NSE stocks")
print("Type 'quit' to exit\n")
while True:
try:
# Get user input
user_input = input("Enter stock symbol: ").strip().upper()
if user_input.lower() == 'quit':
print("๐ Thank you for using Stock Sentiment Analyzer!")
break
if not user_input:
print("โ Please enter a valid stock symbol")
continue
# Get company name (optional)
company_name = input("Enter company name (optional, press Enter to skip): ").strip()
# Perform analysis
analysis = analyzer.get_comprehensive_analysis(user_input, company_name if company_name else None)
# Display results
if analysis:
analyzer.display_analysis(analysis)
except Exception as e:
print(f"โ Error: {str(e)}")
print("Please try again with a different stock symbol")
if __name__ == "__main__":
main() |