import history import functions import matplotlib.pyplot as plt import numpy as np # List of symbols symbols = ['AAPL'] # Create your data frame df = history.get_csv_data(symbols, '01/01/2007', '01/01/2017') # Fill all null values functions.fill_null_values(df) # Get daily returns daily_returns = functions.compute_daily_returns(df) # Generate histogram daily_returns.hist(bins=20) mean_spy = daily_returns['SPY'].mean() std_spy = daily_returns['SPY'].std() plt.axvline(mean_spy, color="w", linestyle="dashed", linewidth=2) plt.axvline(std_spy, color="r", linestyle="dashed", linewidth=2) plt.axvline(-std_spy, color="r", linestyle="dashed", linewidth=2) # mean_aapl = daily_returns['AAPL'].mean() # std_aapl = daily_returns['AAPL'].std() # # plt.axvline(mean_aapl, color='w', linestyle='dashed', linewidth=2) # plt.axvline(std_aapl, color='r', linestyle='dashed',linewidth=2) # plt.axvline(-std_aapl, color='r', linestyle='dashed',linewidth=2) daily_returns['SPY'].hist(bins=20, label='SPY') daily_returns['AAPL'].hist(bins=20, label='AAPL')
import pandas as pd
import matplotlib.pyplot as plt
import scipy.optimize as spo
import numpy as np
import history
df = history.get_csv_data(['AAPL'])
# df['timestamp'] = pd.date_range('2000-1-1', periods=200, freq='D')
print df
import history import functions import numpy as np import math start_date = "2009-01-01" end_date = "2012-12-31" symbols = ['SPY', 'XOM', 'GOOG', 'GLD'] start_value = 1000000 allocations = np.array([0.4, 0.4, 0.1, 0.1]) df = history.get_csv_data(symbols, start_date, end_date) daily_returns = functions.compute_daily_returns(df) normed = functions.compute_normed(df) allocated = functions.position_values(normed, allocations) position_values = np.array(allocated) portfolio_values = position_values.sum(axis=1) daily_returns = daily_returns[1:] cumulative_returns = (portfolio_values[-1] / portfolio_values[1]) - 1 avg_daily_returns = daily_returns.mean() std_daily_returns = daily_returns.std() k = math.sqrt(252) sharpe_ratio = k * avg_daily_returns / std_daily_returns print sharpe_ratio
import history as h
import functions
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
start_date = "2009-01-01"
end_date = "2012-12-31"
dates = pd.date_range(start_date, end_date)
symbols = ['SPY', 'XOM', 'GLD']
df = h.get_csv_data(symbols, start_date, end_date)
daily_returns = functions.compute_daily_returns(df)
# IF the beta value is greater that means that that stock is more reactive to the market.
# Scatter plot SPY vs XOM
daily_returns.plot(kind='scatter', x='SPY', y='XOM')
beta, alpha = np.polyfit(daily_returns['SPY'], daily_returns['XOM'], 1)
plt.plot(daily_returns['SPY'],
beta * daily_returns['SPY'] + alpha,
'-',
color='r')
print "XOM Alpha", alpha
print "XOM Beta", beta
plt.show()
# Scatter plot SPY vs GLD
daily_returns.plot(kind='scatter', x='SPY', y='GLD')
beta, alpha = np.polyfit(daily_returns['SPY'], daily_returns['GLD'], 1)
plt.plot(daily_returns['SPY'],
beta * daily_returns['SPY'] + alpha,
'-',