app.py

import gradio as gr
import pandas as pd
import yfinance as yf
from datetime import datetime, timedelta
import requests
from bs4 import BeautifulSoup
from pattern_finder import score_downward_trend, score_candle, calculate_risk_reward
import urllib3
from datetime import datetime, timedelta
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)


def load_sp500_tickers():
    """Load S&P 500 tickers from Wikipedia."""
    url = "https://en.wikipedia.org/wiki/List_of_S%26P_500_companies"
    response = requests.get(url, verify=False)
    soup = BeautifulSoup(response.content, 'html.parser')
    table = soup.find('table', {'id': 'constituents'})
    tickers = []
    if table:
        for row in table.find_all('tr')[1:]:
            cells = row.find_all('td')
            if cells:
                ticker = cells[0].text.strip()
                tickers.append(ticker)
    return tickers



def load_data(ticker, interval="1d"):
    """Load stock data using yfinance with a specified interval."""
    end_date = datetime.today()
    start_date = end_date - timedelta(days=365)  # Get 1 year of data
    data = yf.download(ticker, start=start_date, end=end_date, interval=interval)
    return data



def calculate_sma(data, window):
    """Calculate the Simple Moving Average (SMA) for a given window."""
    return data['Close'].rolling(window=window).mean()

def calculate_ema(data, window):
    """Calculate the Exponential Moving Average (EMA) for a given window."""
    return data['Close'].ewm(span=window, adjust=False).mean()

def average_downtrend(data, method, window=4):
    """Calculate the average difference between consecutive prices for the last 'window' candles."""
    if len(data) < window:
        return 0.0
    price_diffs = data[method].diff().iloc[-window:]
    avg_diff = price_diffs.mean()
    return avg_diff if avg_diff < 0 else 0.0


def score_candle(candle, prev_candle, trend_strength):
    """Score a single candle based on its characteristics and previous candle."""
    open_price = candle['Open']
    close_price = candle['Close']
    low_price = candle['Low']
    high_price = candle['High']
    prev_close = prev_candle['Close']

    # Bottom and top wick lengths
    bottom_wick_length = min(open_price, close_price) - low_price
    top_wick_length = high_price - max(open_price, close_price)

    # Initial score based on trend strength
    score = trend_strength * 2

    # Doji: Open and Close are almost the same (small body)
    if abs(open_price - close_price) <= 0.1 * (high_price - low_price):  # Adjust tolerance if needed
        score += 5  # Bonus points for doji candles

    # Hammer: Small body at the top, long bottom wick (typical reversal candle)
    if close_price < open_price and bottom_wick_length > 2 * (open_price - close_price):
        score += 7  # Extra points for hammer-like candles

    # Bottom Tailing Wick: Long bottom wick compared to the overall range
    if bottom_wick_length > 0.5 * (high_price - low_price):
        score += 6  # Extra points for bottom tailing wick

    # Additional logic: Boost red candles with long bottom wicks following a downtrend
    if close_price < open_price and bottom_wick_length > 0.5 * (open_price - close_price):
        score += 3  # Boost for red candle with long bottom wick

    # Penalize if the current close is higher than the previous close
    if close_price > prev_close:
        score -= ((close_price - prev_close) / prev_close) * 100


    return score

def score_today_candle(data, window=4):
    """Score today's candle based on the downtrend from the past 'window' days."""
    if len(data) < window + 1:
        return 0  # Not enough data

    today_candle = data.iloc[-1]
    prev_candle = data.iloc[-2]
    
    close_price = today_candle['Close']


    previous_data = data.iloc[-(window+1):-1]
    down_High = average_downtrend(previous_data, method="High",window=window) + average_downtrend(previous_data, method="High",window=7) / 2
    down_Close = average_downtrend(previous_data, method="Close",window=window) + average_downtrend(previous_data, method="Close",window=7) / 2


    avg_downtrend = (down_High + down_Close) / 2
  
    if avg_downtrend == 0.0:
        return -1

    # Calculate SMAs for the last row
    sma_50 = calculate_sma(data, window=50).iloc[-1]
    sma_200 = calculate_sma(data, window=200).iloc[-1]
    sma_20 = calculate_sma(data, window=20).iloc[-1]
    
    ema_10 = calculate_ema(data, window=10).iloc[-1]
    
    if (close_price < ema_10) or (close_price < sma_20) or  (close_price < sma_50) or (close_price < sma_200):
        return -1

    
    return score_candle(today_candle, prev_candle, abs(avg_downtrend))

def scan_sp500(top_n=25, interval="1d", progress=gr.Progress()):
    tickers = load_sp500_tickers()
    scores = []
    tickers.append("QQQ")

    for i, ticker in enumerate(progress.tqdm(tickers)):
        data = load_data(ticker, interval)
        if not data.empty:
            score = score_today_candle(data)
            if score > 0:
                scores.append((ticker, score))
    
    scores = sorted(scores, key=lambda x: x[1], reverse=True)
    return scores[:top_n]

def next_business_day(date):
    next_day = date + timedelta(days=1)
    while next_day.weekday() >= 5:  # 5 = Saturday, 6 = Sunday
        next_day += timedelta(days=1)
    return next_day



def gradio_scan_sp500(top_n, interval, progress=gr.Progress()):
    progress(0, desc="Downloading Data")
    tickers = load_sp500_tickers()
    tickers.append("QQQ")
    
    progress(0.3, desc="Running Scanner")
    results = scan_sp500(top_n, interval, progress)
    
    # Get the last date of the data and find the next business day
    last_data = load_data(results[0][0], interval)  # Load data for the first ticker in results
    last_date = last_data.index[-1].date()
    next_market_day = next_business_day(last_date)
    date_created = next_market_day.strftime("%Y-%m-%d")
    
    output = f"Scan Results for Market Open on: {date_created}\n\n"
    output += "Top {} stocks based on pattern finder score:\n\n".format(top_n)
    for ticker, score in results:
        output += "{}: Total Score = {:.2f}\n".format(ticker, score)
    return output

iface = gr.Interface(
    fn=gradio_scan_sp500,
    inputs=[
        gr.Slider(minimum=1, maximum=100, step=1, label="Number of top stocks to display", value=25),
        gr.Dropdown(choices=["1m", "1d", "1wk", "1mo"], label="Data Interval", value="1d"),
    ],
    outputs="text",
    title="S&P 500 Stock Scanner",
    description="Scan S&P 500 stocks and display top N stocks based on today's candle score.",
    allow_flagging="never",
)


if __name__ == "__main__":
    iface.launch(server_name="0.0.0.0", server_port=7860)