def Beta_graph(ticker, start_date, end_date):
if start_date is not None and end_date is not None:
stock = func.getPrices(collection_price, ticker, start_date, end_date,
"close")
stock = stock.drop_duplicates()
stock = stock['close'].pct_change()[1:].values
SPY = func.getPrices(collection_price, "SPY", start_date, end_date,
"close")
SPY = SPY.drop_duplicates()
SPY = SPY['close'].pct_change()[1:].values
try:
p = func.linreg(stock, SPY)
pts = go.Scatter(x=SPY, y=stock, mode='markers', showlegend=False)
line = go.Scatter(x=SPY,
y=p[0] + p[1] * SPY,
showlegend=False,
mode='lines')
data = [pts, line]
layout = dict(xaxis=dict(zeroline=False, linewidth=1, mirror=True),
yaxis=dict(zeroline=False, linewidth=1, mirror=True),
title='Beta to SPY')
fig = dict(data=data, layout=layout)
except:
fig = {}
else:
fig = {}
return fig
def returnDist_graph(start_date, end_date, accountType):
if start_date is not None and end_date is not None:
if accountType == 'paperAccount':
df = func.getAccoutValue(collection_account, "NetLiquidation")
elif accountType == 'liveAccount':
df = func.getAccoutValue(collection_account_real, "NetLiquidation")
df = df[df['date'] >= start_date]
spy = func.getPrices(collection_price, 'SPY', start_date, end_date,
'close')
spy = spy.drop_duplicates()
spy = spy['close'].pct_change()[1:]
violin = dict(type='violin',
y=pd.to_numeric(df['value']).pct_change()[1:],
box=dict(visible=True),
meanline=dict(visible=True),
name='Portfolio')
violin_spy = dict(type='violin',
y=spy,
box=dict(visible=True),
meanline=dict(visible=True),
name='SPY')
data = [violin, violin_spy]
layout = dict(xaxis=dict(zeroline=False, linewidth=1, mirror=True),
yaxis=dict(zeroline=False, linewidth=1, mirror=True),
title='Violin Plot of Return')
fig = dict(data=data, layout=layout)
else:
fig = {}
return fig
def update_correlation_map(start_date, end_date):
df = func.getPassedPortfolio(collection_portfolio)
result = pd.DataFrame()
for index, row in df.iterrows():
lst = {}
symbol = row['Ticker']
price = func.getPrices(collection_price, symbol, start_date, end_date,
'close')
price = price.drop_duplicates()
try:
lst[symbol] = price['close'].tolist()
result = pd.concat([result, pd.DataFrame(lst)], axis=1)
except:
result = pd.DataFrame(lst)
cor = result.corr()
arr = cor.values
trace = go.Heatmap(z=arr, x=cor.columns, y=cor.columns)
layout = dict(height=800, title='Stock Correlation')
data = [trace]
fig = dict(data=data, layout=layout)
return fig
def QQ_graph(ticker, start_date, end_date):
if start_date is not None and end_date is not None:
X_lognorm = func.getPrices(collection_price, ticker, start_date,
end_date, "close")
X_lognorm = X_lognorm['close'].tolist()
qq = stats.probplot(X_lognorm, dist='norm', sparams=(1))
x = np.array([qq[0][0][0], qq[0][0][-1]])
pts = go.Scatter(x=qq[0][0],
y=qq[0][1],
mode='markers',
showlegend=False)
line = go.Scatter(x=x,
y=qq[1][1] + qq[1][0] * x,
showlegend=False,
mode='lines')
data = [pts, line]
layout = dict(xaxis=dict(zeroline=False, linewidth=1, mirror=True),
yaxis=dict(zeroline=False, linewidth=1, mirror=True),
title='QQ Plot')
fig = dict(data=data, layout=layout)
else:
fig = {}
return fig
def equity_graph(start_date, end_date, accountType):
if start_date is not None and end_date is not None:
if accountType == 'paperAccount':
df = func.getAccoutValue(collection_account, "NetLiquidation")
elif accountType == 'liveAccount':
df = func.getAccoutValue(collection_account_real, "NetLiquidation")
df = df[df['date'] >= start_date]
df_2 = func.getPrices(collection_price, 'SPY', start_date, end_date,
'close')
amount = float(df.iloc[0]['value'])
price = float(df_2.iloc[0]['close'])
shares = amount / price
#extract the date only, and row back one day becoz account value is today, price is yesterday
line = go.Scatter(x=df['date'].apply(
lambda x: x.date() - timedelta(days=1)).tolist(),
y=df['value'],
showlegend=True,
mode='lines',
name='Equity Curve')
line_spy = go.Scatter(x=df_2['date'],
y=df_2['close'] * shares,
showlegend=True,
mode='lines',
name='SPY')
data = [line, line_spy]
layout = dict(xaxis=dict(zeroline=False, linewidth=1, mirror=True),
yaxis=dict(
zeroline=False,
linewidth=1,
mirror=True,
type='log',
),
title='Equity Curve')
fig = dict(data=data, layout=layout)
else:
fig = {}
return fig
def returnVSSPY_graph(start_date, end_date, accountType):
if start_date is not None and end_date is not None:
if accountType == 'paperAccount':
df = func.getAccoutValue(collection_account, "NetLiquidation")
elif accountType == 'liveAccount':
df = func.getAccoutValue(collection_account_real, "NetLiquidation")
df = df[df['date'] >= start_date]
"""
start_date = df['date'].tolist()
start_date.sort()
start_date = start_date[-5].strftime('%Y-%m-%d %H:%M:%S.%f')
start_date = start_date.split(" ")[0]
#print(start_date)
"""
spy = func.getPrices(collection_price, 'SPY', start_date, end_date,
'close')
spy = spy.drop_duplicates()
#extract the date only, and row back one day becoz account value is today, price is yesterday
bar = go.Bar(
x=df['date'].apply(
lambda x: x.date() - timedelta(days=1)).tolist()[1:],
y=pd.to_numeric(df['value']).pct_change()[1:],
name='Equity',
)
bar_spy = go.Bar(x=spy['date'].tolist()[1:],
y=spy['close'].pct_change()[1:],
name='SPY')
data = [bar, bar_spy]
layout = dict(title='Return V.S SPY')
fig = dict(data=data, layout=layout)
else:
fig = {}
return fig