def full_summary(self):
return {
'csv': {
'pnl_csv': da.csv(self.pnl, True),
'cum_pnl_csv': da.csv(self.pnl.cumsum(), True)
},
'cps': self.cps()
}
def __init__(self, ticker, json):
self.ticker = ticker
self.df = da.dataframe(ticker)
self.df_spy = da.dataframe('SPY')
self.price_csv = da.csv(self.df['Adj_Close'], json)
pnl = self.df['Adj_Close'].diff()
pnl[0] = 0
self.pnl_csv = da.csv(pnl, json)
self.json = json
def __init__(self, ticker, json):
self.ticker = ticker
self.df = da.dataframe(ticker)
self.df_spy = da.dataframe('SPY')
self.price_csv = da.csv(self.df['Adj_Close'], json)
pnl = self.df['Adj_Close'].diff()
pnl[0] = 0
self.pnl_csv = da.csv(pnl, json)
self.json = json
def __init__(self, ticker1, ticker2, json):
self.df1 = da.dataframe(ticker1)
self.df2 = da.dataframe(ticker2)
self.ticker1 = ticker1
self.ticker2 = ticker2
self.price_csv1 = da.csv(self.df1['Adj_Close'], json)
self.price_csv2 = da.csv(self.df2['Adj_Close'], json)
pnl1 = self.df1['Adj_Close'].diff()
pnl1[0] = 0
pnl2 = self.df2['Adj_Close'].diff()
pnl2[0] = 0
self.pnl_csv1 = da.csv(pnl1, json)
self.pnl_csv2 = da.csv(pnl2, json)
self.json = json
self.flip = 'False'
if self.df2.index[0] > self.df1.index[0]:
self.flip = 'True'
def __init__(self, ticker1, ticker2, json):
self.df1 = da.dataframe(ticker1)
self.df2 = da.dataframe(ticker2)
self.ticker1 = ticker1
self.ticker2 = ticker2
self.price_csv1 = da.csv(self.df1["Adj_Close"], json)
self.price_csv2 = da.csv(self.df2["Adj_Close"], json)
pnl1 = self.df1["Adj_Close"].diff()
pnl1[0] = 0
pnl2 = self.df2["Adj_Close"].diff()
pnl2[0] = 0
self.pnl_csv1 = da.csv(pnl1, json)
self.pnl_csv2 = da.csv(pnl2, json)
self.json = json
self.flip = "False"
if self.df2.index[0] > self.df1.index[0]:
self.flip = "True"
def cointegration_test(self, start, end):
x = da.selection(self.df1["Adj_Close"], start, end)
y = da.selection(self.df2["Adj_Close"], start, end)
ols_result = sm.OLS(x, y).fit()
resid = ols_result.resid
# compute regression residuals
adfuller_results = ts.adfuller(resid)
print "cointegration: " + str(adfuller_results[1])
bound = pd.Series(movvar(resid, 30)) ** 0.5
resid_rolling_average = da.exp_moving_average(resid, 0.1)
resid_with_bounds = np.array([resid_rolling_average, bound, -bound]).T
columns = ["Residuals", "Upper", "Lower"]
resid_with_bounds_df = pd.DataFrame(resid_with_bounds, index=resid.index, columns=columns)
return {"csv": da.csv(resid_with_bounds_df, True)}
def cointegration_test(self, start, end):
x = da.selection(self.df1['Adj_Close'], start, end)
y = da.selection(self.df2['Adj_Close'], start, end)
ols_result = sm.OLS(x, y).fit()
resid = ols_result.resid
# compute regression residuals
adfuller_results = ts.adfuller(resid)
print "cointegration: " + str(adfuller_results[1])
bound = pd.Series(movvar(resid, 30))**0.5
resid_rolling_average = da.exp_moving_average(resid, 0.1)
resid_with_bounds = np.array([resid_rolling_average, bound, -bound]).T
columns = ['Residuals', 'Upper', 'Lower']
resid_with_bounds_df = pd.DataFrame(resid_with_bounds,
index=resid.index,
columns=columns)
return {'csv': da.csv(resid_with_bounds_df, True)}
def full_summary(self):
return {'csv' : {'pnl_csv' : da.csv(self.pnl, True), 'cum_pnl_csv' : da.csv(self.pnl.cumsum(), True)},
'cps' : self.cps()}
