import pandas as pd
import numpy as np
import yfinance as yf
import warnings
warnings.filterwarnings('ignore')

# Beginner Example 1: Load stock prices for multiple tickers
tickers = ['AAPL', 'MSFT', 'GOOGL', 'AMZN', 'TSLA']
df = yf.download(tickers, period='1y', auto_adjust=True, progress=False)
df = df['Close'].reset_index()
df.columns.name = None
print(df.shape)
print(df.head(3))

# Beginner Example 2: Load daily returns for the SPY ETF
data = yf.download('SPY', period='2y', auto_adjust=True, progress=False)
spy_df = data[['Close','Volume']].reset_index()
spy_df.columns = ['date','price','volume']
spy_df['daily_return'] = (spy_df['price'].pct_change() * 100).round(4)
spy_df = spy_df.dropna().reset_index(drop=True)
print(spy_df.shape)
print(spy_df.head(3))

# Beginner Example 3: Load multi-ticker OHLCV data in long format
tickers = ['AAPL','MSFT','GOOGL','AMZN','TSLA']
ohlcv = yf.download(tickers, period='1y', auto_adjust=True, progress=False)
ohlcv = ohlcv.stack(future_stack=True).rename_axis(['Date','Ticker']).reset_index()
ohlcv.columns.name = None
print(ohlcv.shape)
print(ohlcv.head(3))

# Beginner Example 4: Select pricing data for a single ticker from long format
aapl_prices = ohlcv[ohlcv['Ticker'] == 'AAPL']
print(aapl_prices.shape)
print(aapl_prices.head(3))

# Beginner Example 5: Convert long to wide format for close prices
wide_prices = ohlcv.pivot(index='Date', columns='Ticker', values='Close')
print(wide_prices.shape)
print(wide_prices.head(3))

# Intermediate Example 1: Plot the closing prices of all five stocks
import matplotlib.pyplot as plt
plt.figure(figsize=(12,6))
for tkr in tickers:
    plt.plot(wide_prices.index, wide_prices[tkr], label=tkr)
plt.title('Close Prices of Major Tech Stocks (1Y)')
plt.xlabel('Date')
plt.ylabel('Price (USD)')
plt.legend()
plt.show()

# Intermediate Example 2: Compute daily returns for all tickers (wide format)
returns = wide_prices.pct_change().dropna() * 100
returns = returns.round(2)
print(returns.head(3))

# Intermediate Example 3: Compute rolling 20-day volatility (standard deviation)
vol_20d = returns.rolling(window=20).std().round(2)
print(vol_20d.tail(3))

# Intermediate Example 4: Correlation between the daily returns of stocks
corr_matrix = returns.corr()
print(corr_matrix)

# Intermediate Example 5: Detect missing data and handle it
print(ohlcv.isnull().sum())
ohlcv_clean = ohlcv.dropna()
print('Data after dropping missing rows:', ohlcv_clean.shape)

# Advanced Example 1: Resample daily returns to monthly returns
monthly_returns = wide_prices.resample('M').last().pct_change().dropna() * 100
monthly_returns = monthly_returns.round(2)
print(monthly_returns.head(3))

# Advanced Example 2: Compute 50-day moving averages as a trend indicator
ma_50 = wide_prices.rolling(window=50).mean()
print(ma_50.tail(3))

# Advanced Example 3: Find the best/worst single-day return for each ticker
best_returns = returns.max()
worst_returns = returns.min()
print('Best single-day returns (%):')
print(best_returns)
print('Worst single-day returns (%):')
print(worst_returns)

# Advanced Example 4: Create trading signals using SMA crossover for AAPL
data = yf.download('AAPL', period='1y', auto_adjust=True, progress=False)
df_sma = data[['Close','Volume']].reset_index()
df_sma.columns = ['date','close','volume']
df_sma['sma20'] = df_sma['close'].rolling(20).mean().round(2)
df_sma['sma50'] = df_sma['close'].rolling(50).mean().round(2)
df_sma['signal'] = np.where(df_sma['sma20'] > df_sma['sma50'], 1, -1)
print(df_sma.tail(3))

# Error Example: Wrong column name in resample
try:
    _ = spy_df.resample('M').last()['close']
except Exception as e:
    print('Error:', e)

# Error Example: Handling NA values by forward filling
aapl_prices_ffill = aapl_prices.fillna(method='ffill')
print(aapl_prices_ffill.isnull().sum())

# End-to-end Problem: Identify the highest single-day loss for TSLA in the past year and plot it
np.random.seed(42)  # for reproducibility
tsla_prices = ohlcv[ohlcv['Ticker'] == 'TSLA'].copy()
tsla_prices['Return'] = tsla_prices['Close'].pct_change() * 100
worst_day = tsla_prices['Return'].idxmin()
print('Worst day for TSLA:', tsla_prices.loc[worst_day, 'Date'], 'Return:', round(tsla_prices.loc[worst_day, 'Return'],2), '%')
import matplotlib.pyplot as plt
plt.figure(figsize=(10, 5))
plt.plot(tsla_prices['Date'], tsla_prices['Close'], label='TSLA Close')
plt.scatter(tsla_prices.loc[worst_day, 'Date'], tsla_prices.loc[worst_day, 'Close'], color='red', label='Worst Day')
plt.title('TSLA Closing Prices with Worst Single-Day Loss Highlighted')
plt.xlabel('Date')
plt.ylabel('Price (USD)')
plt.legend()
plt.show()