def update_candle(n, tick, start_date, end_date):
code = 'WSE/' + tick
df = pdr.get_data_quandl(code,
api_key='fHsXs9kzqak6UF1haCww',
start=start_date,
end=end_date)
figure = {
'data': [
go.Candlestick(x=df.index,
open=df['Open'],
close=df['Close'],
low=df['Low'],
high=df['High'],
increasing_line_color='lime',
decreasing_line_color='red')
],
'layout':
go.Layout(height=350,
title='Candlestick chart for {}'.format(tick),
plot_bgcolor='#434247',
paper_bgcolor='#434247',
yaxis={'gridcolor': '#35353b'},
xaxis={'gridcolor': '#35353b'},
font={'color': '#d8d8d8'},
margin=dict(t=30, b=0, l=60, r=60))
}
return figure
def update_sharpe(n, ticks, tries, start_date, end_date):
if len(ticks) >= 2:
df = pd.DataFrame()
ret_arr = np.zeros(tries)
vol_arr = np.zeros(tries)
sr_arr = np.zeros(tries)
for name in ticks:
tick = 'WSE/' + name
df_tick = pdr.get_data_quandl(tick,
api_key='fHsXs9kzqak6UF1haCww',
start=start_date,
end=end_date)
df_tick.rename({'Close': name}, axis=1, inplace=True)
df = df.join(df_tick[name], how='outer')
logs = np.log(df / df.shift(-1))
all_weights = np.zeros((tries, len(logs.columns)))
for i in range(tries):
weights = np.array(np.random.random(len(logs.columns)))
weights = weights / np.sum(weights)
all_weights[i, :] = weights
ret_arr[i] = np.sum(logs.mean() * weights * len(logs))
vol_arr[i] = np.sqrt(
np.dot(weights.T, np.dot(logs.cov() * len(logs), weights)))
sr_arr[i] = ret_arr[i] / vol_arr[i]
weights_as_string = []
for weight in all_weights:
weights_as_string.append(', '.join(
str(round(i, 3)) for i in weight))
df = pd.DataFrame({
'returns': ret_arr,
'volatility': vol_arr,
'sharpe': sr_arr,
'weights': weights_as_string
})
figure = px.scatter(df,
x='volatility',
y='returns',
color='sharpe',
color_continuous_scale='RdPu_r',
hover_data=['weights'])
figure.update_traces(marker=dict(size=9, line=dict(width=1)))
figure.update_layout(title='Portfolio allocation for {}'.format(
', '.join(ticks)),
yaxis={'gridcolor': '#35353b'},
xaxis={'gridcolor': '#35353b'},
xaxis_title='Volatility',
yaxis_title='Returns',
plot_bgcolor='#434247',
paper_bgcolor='#434247',
margin=dict(t=40, b=30, l=60, r=60),
font={'color': '#d8d8d8'})
return figure
def update_line(n, ticks, start_date, end_date):
traces = []
for name, color in zip(ticks, colors):
tick = 'WSE/' + name
df = pdr.get_data_quandl(tick,
api_key='fHsXs9kzqak6UF1haCww',
start=start_date,
end=end_date)
traces.append({
'x': df.index,
'y': df['Close'],
'name': name,
'mode': 'lines',
'line': dict(color=color)
})
figure = {
'data': traces,
'layout': {
'height': 300,
'title': 'Stock chart for {}'.format(', '.join(ticks)),
'yaxis': {
'gridcolor': '#35353b'
},
'xaxis': {
'gridcolor': '#35353b'
},
'plot_bgcolor': '#434247',
'paper_bgcolor': '#434247',
'margin': dict(t=40, b=30, l=60, r=60),
'font': {
'color': '#d8d8d8'
}
}
}
return figure
def download_stock(symbol, start_date, end_date, save=True):
stock = pdr.get_data_quandl(symbol, start_date, end_date)
stock = stock[::-1]
stock.columns = [s.lower() for s in stock.columns]
stock = stock[['open', 'high', 'low', 'close', 'volume']]
if save:
name = symbol + '.csv'
path = config.path_to_data + '/' + name
stock.to_csv(path)
return stock
def create_from_quandl(cls, instrument, *args, **kwargs):
""" Загрузка данных через pandas_datareader is https://www.quandl.com
"""
df = pdr.get_data_quandl(instrument,
api_key=os.environ.get('QUANDL_API_KEY'),
*args,
**kwargs)
return cls.create_(
instrument, df.index.values, df, **{
'o': 'AdjOpen',
'h': 'AdjHigh',
'l': 'AdjLow',
'c': 'AdjClose',
'v': 'AdjVolume'
})
def update_indices(n):
rates = []
changes = []
traces = []
try:
req = requests.get("https://stooq.pl/t/?i=528")
soup = BeautifulSoup(req.text, 'html.parser')
for index in indices:
rates.append(
float(soup.find(text=index).next_element.next_sibling.text))
try:
changes.append(
float(
soup.find(text=index).next_element.next_sibling.
next_sibling.next_sibling.text))
except:
continue
if len(changes) == len(indices):
for number, index, rate, change in zip(range(len(indices)),
indices, rates, changes):
traces.append(
go.Indicator(mode='number+delta',
title=index,
title_font_size=14,
number_font_size=20,
value=rate,
delta={
'reference': rate - change,
'relative': True,
'valueformat': '.2%'
},
domain={
'row': 1,
'column': number
}))
else:
for number, index, rate in zip(range(len(indices)), indices,
rates):
traces.append(
go.Indicator(mode='number',
title=index,
title_font_size=14,
number_font_size=20,
value=rate,
domain={
'row': 1,
'column': number
}))
figure = {
'data':
traces,
'layout':
go.Layout(grid={
'rows': 1,
'columns': len(indices),
'pattern': "independent"
},
margin=dict(t=30, b=0, l=0, r=0),
plot_bgcolor='#35353b',
paper_bgcolor='#35353b',
font={'color': '#d8d8d8'},
height=70)
}
return figure
except:
try:
inds = [
'WIG', 'WIG20', 'MWIG40', 'SWIG80', 'WIG_ODZIEZ', 'WIG_LEKI'
]
traces = []
for index, number in zip(inds, range(len(inds))):
code = 'WSE/' + index
df = pdr.get_data_quandl(code,
api_key='fHsXs9kzqak6UF1haCww',
start=datetime.today().date() -
timedelta(days=5))
traces.append(
go.Indicator(mode='number',
title=index,
title_font_size=14,
number_font_size=20,
value=df['Close'][0],
domain={
'row': 1,
'column': number
}))
figure = {
'data':
traces,
'layout':
go.Layout(grid={
'rows': 1,
'columns': len(inds),
'pattern': "independent"
},
margin=dict(t=30, b=0, l=0, r=0),
plot_bgcolor='#35353b',
paper_bgcolor='#35353b',
font={'color': '#d8d8d8'},
height=70)
}
return figure
except:
figure = {
'data': [
go.Indicator(
title='Could not connect to stooq to get index data',
title_font_size=20,
number_font_size=1)
],
'layout':
go.Layout(plot_bgcolor='#35353b',
paper_bgcolor='#35353b',
font={'color': 'tomato'},
height=70)
}
return figure
def get_stock_quandl(stock):
return DataFrame(
pdr.get_data_quandl(stock,
api_key="A83E-dJAdqmse9P-Tpng",
session=session))
#import what tickers were being monitored on twitter
tickers = open('TickersToTrack.txt').read().splitlines()
#plot and to some analysis
for tick in tickers:
s = tick.lower()
data = df[df.Stock == s]
#plt.plot(data['Date'],data['Sentiment'],'-r')
#plt.title(s)
#plt.xlabel('Date')
#plt.ylabel('Sentiment')
#plot tickers with stock price
startDate = datetime.datetime(2018, 2, 1)
endDate = datetime.datetime(2018, 2,
5) #must be datetime.datetime(yyyy,mm,dd)
# Import stock data using pandasreader
try:
dfstock = pdr.get_data_quandl(tick[1::], start=startDate, end=endDate)
dfstock = dfstock.reindex(
index=dfstock.index[::-1]) #flip index so 0th row is oldest date
plt.plot(dfstock['Close'], data['Sentiment'], 'g-')
plt.title('Close to Sentiment: ' + tick[1::])
plt.xlabel('Close')
plt.ylabel('Sentiment')
plt.show()
except:
print("Couldn't get data for ticker: " + tick[1::])
# scatter plot the data
import pandas_datareader.data as web
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from scipy.optimize import minimize
start = '2017-04-22'
end = '2018-04-22'
from pandas_datareader import data, wb
import pandas_datareader as pdr
pdr.get_data_yahoo(
'AAPL', start=start,
end=end) # if by default, returns 2009-2019, 2403 rows, daily stock data
pdr.get_data_quandl('EOD/BA',
start=start,
end=end,
api_key='DGCrDAbR7GMbx7v9wSCW')
def get_risk(prices):
return (prices / prices.shift(1) - 1).dropna().std().values
def get_return(prices):
return ((prices / prices.shift(1) - 1).dropna().mean() *
np.sqrt(250)).values
# symbols = ['BA', 'AAPL', 'BRK_A', 'MSFT']
symbols = ['BA', 'C', 'AAL', 'NFLX']
prices = pd.DataFrame(index=pd.date_range(start, end))
def data(ticker): return (pdr.get_data_quandl(ticker, start=startdate, end=enddate))
#pd.scatter_matrix(daily_pct_change, diagonal='kde', alpha=0.1,figsize=(12,12)) # this function is broken, use pd.plotting as pdp
pdp.scatter_matrix(daily_pct_change, diagonal='kde', alpha=0.1,figsize=(12,12))
plt.savefig('all_data_quandl_4_scattermatrix.png')
plt.show()
'''
Moving Windows, when u compute the stats on window of data by a particular period of time, then slide the window
across the data by a specific interval. Continually calculated as long as the window falls first within the dates
of the time series.
'''
#data = quandl.get('AAPL', authtoken='-VQTkx89BPNGVyjKvD1p')
# from earlier reference (file1)
aapl = pdr.get_data_quandl('AAPL',
start=datetime.datetime(2006, 10, 1),
end=datetime.datetime(2018, 1, 1))
#must use own data, can request 20 pulls per 10 min from quandl (uses AdjClose)
#csv (uses Close)
#aapl = pd.read_csv('aapl_ohlc.csv')
adj_close_px_aapl = aapl['AdjClose']
# Calculate the moving average:
# rolling_mean(window[, min_periods, ...]) 'Moving mean'.
# rolling_std(arg, window[, min_periods, ...]) 'Moving standard deviation'.
# rolling_max(arg, window[, min_periods, ...]) 'Moving max of 1d array of dtype=float64 along axis=0 ignoring NaNs'.
# rolling_var(arg, window[, min_periods, ...]) 'Numerically stable implementation using Welford’s method'.
import pandas as pd
import pandas_datareader as pdr
pd.core.common.is_list_like = pd.api.types.is_list_like
# 데이터프레임 저장
# www.quandl.com에 가입하여 api_key받기
tesla = pdr.get_data_quandl('TSLA', api_key='받은 키 입력~')
# dataframe을 파일로 저장
tesla.to_csv('tesla_stock_quandl.csv')
# date열을 datetime형으로 변환
tesla = pd.read_csv('tesla_stock_quandl.csv', parse_dates=[0])
# date열을 tesla dataframe의 인덱스로 저장
tesla.index = tesla['Date']
# 예시로 2015년의 데이터 추출
print(tesla['2015'].iloc[:5, :5])
# data열의 최솟값을 빼서 데이터 수집이후 시간흘렀는지 확인해보기
# 그 시간을 ref_date열로 추가
tesla['ref_date'] = tesla['Date'] - tesla['Date'].min()
# ref_date열을 인덱스로 지정해서 최초 5일의 데이터 추출
tesla.index = tesla['ref_date']
print(tesla['5 days':].iloc[:5, :5])