#!/usr/bin/env python
# coding: utf-8
# In[2]:
import pandas as pd
import datetime
import pandas_datareader.data as web
from pandas import Series, DataFrame
start = datetime.datetime(2010, 1, 1)
end = datetime.datetime(2020, 6, 4)
df = web.DataReader("AAPL", 'yahoo', start, end)
df.tail()
# In[3]:
close_px = df['Adj Close']
mavg = close_px.rolling(window=100).mean()
# In[4]:
mavg
def test_iex_bad_symbol_list(self):
with pytest.raises(Exception):
web.DataReader(["AAPL", "BADTICKER"], "iex", self.start, self.end)
def test_single_symbol(self):
df = web.DataReader("AAPL", "iex", self.start, self.end)
assert list(df) == ["open", "high", "low", "close", "volume"]
assert len(df) == 476
cur = conn.cursor(cursor_factory=RealDictCursor)
cur.execute("""
SELECT * from mock_watchlist
ORDER BY ticker ASC
;
""")
res = cur.fetchall()
tickers = [x['ticker'] for x in res]
# currently have to tickers (BHP, CBA)
print("Tickers from db: {}".format(tickers))
# for now, just query one stock maybe NAB?
# make dataframe (build custom function that will take ticker, and start and end, and return the DF)
for tick in tickers:
# for now just get 30 days
end = datetime.date.today()
start = end - datetime.timedelta(days=30)
df = data.DataReader('{}.AX'.format(tick),
start=start,
end=end,
data_source='yahoo')[[
'High', 'Low', 'Open', 'Adj Close'
]]
df['Up Today'] = df['Adj Close'] > df['Open']
df['% Change'] = df['Adj Close'].pct_change(periods=1) * 100
# Add moving averages?
# Add momentum indicator?
# Compute here what kind of candlestick it is (maybe only last ten?)
print(df)
# build function that will take a candlestick and define it (ie, doji, etc, use the Bedford book for defintions)
import pandas_datareader.data as web
import datetime
import matplotlib.pyplot as plt
# data
start = datetime.datetime(2010, 1, 1)
end = datetime.datetime(2016, 3, 19)
data = web.DataReader("AAPL", "yahoo", start, end)
plt.plot(data.index, data['Adj Close'])
plt.show()
import datetime as dt
import pandas_datareader.data as web
from machineTime import getMachineData
now = dt.datetime.combine(dt.datetime.now(), dt.time())
date = dt.datetime.fromtimestamp(1588341840)
start = date
end = date + dt.timedelta(days=1) # dt.datetime.now()
start = dt.datetime(2015, 7, 1)
end = dt.datetime(2020, 7, 14)
'''if(end > now):
end = date'''
dfs = []
symbols = [
"NBEV", "NEM", "AMD", "MKTX", "NVDA", "REGN", 'NLOK', 'HUM', 'VRTX', 'RMD',
'APPL', 'ODFL', 'MSCI', 'CTXS', 'DVA', 'SBAC', 'TGT', 'DG', 'MSFT', 'LDOS',
'ANSS'
]
for count, symbol in enumerate(symbols):
try:
dfs.append(web.DataReader(symbol, 'yahoo', start, end))
except:
pass
getMachineData(dfs)
import pandas as pd
import matplotlib.pyplot as plt
import pandas_datareader.data as web
# Get GS Data from Yahoo
gs = web.DataReader("078930.KS", "yahoo", "2014-01-01", "2016-03-06")
new_gs = gs[gs['Volume'] != 0]
# Moving average
ma5 = new_gs['Adj Close'].rolling(window=5).mean()
ma20 = new_gs['Adj Close'].rolling(window=20).mean()
ma60 = new_gs['Adj Close'].rolling(window=60).mean()
ma120 = new_gs['Adj Close'].rolling(window=120).mean()
# Insert columns
new_gs.insert(len(new_gs.columns), "MA5", ma5)
new_gs.insert(len(new_gs.columns), "MA20", ma20)
new_gs.insert(len(new_gs.columns), "MA60", ma60)
new_gs.insert(len(new_gs.columns), "MA120", ma120)
# Plot
plt.plot(new_gs.index, new_gs['Adj Close'], label="Adj Close")
plt.plot(new_gs.index, new_gs['MA5'], label="MA5")
plt.plot(new_gs.index, new_gs['MA20'], label="MA20")
plt.plot(new_gs.index, new_gs['MA60'], label="MA60")
plt.plot(new_gs.index, new_gs['MA120'], label="MA120")
plt.legend(loc='best')
plt.grid()
plt.show()
import pandas as pd
import datetime
# import pandas.io.data as web
# from pandas_datareader import data, wb
import pandas_datareader.data as web
import matplotlib.pyplot as plt
from matplotlib import style
style.use('ggplot')
start = datetime.datetime(2010, 1, 1)
end = datetime.datetime(2015, 1, 1)
df = web.DataReader("XOM", "yahoo", start, end)
print(df.head())
df['Adj Close'].plot()
plt.show()
import datetime as dt
import matplotlib.pyplot as plt
print(
'This program will plot stock prices for you, just follow the prompts as they come...'
)
date_entry = input('Enter the start date for your plot YYYY-MM-DD format: ')
year, month, day = map(int, date_entry.split('-'))
start = dt.datetime(year, month, day)
date_entry2 = input('Enter the end date for your plot YYYY-MM-DD format: ')
year, month, day = map(int, date_entry2.split('-'))
end = dt.datetime(year, month, day)
ticker = input('Enter the ticker you wish to plot: ')
while True:
query = input('Do you want to compare it to the market index? ')
first_l = query[0].lower()
if query == '' or not first_l in ['y', 'n']:
print('Please answer with yes or no!')
else:
break
if first_l == 'y':
df = web.DataReader([ticker, 'SPY'], 'yahoo', start, end)
df['Adj Close'].plot()
plt.show()
if first_l == 'n':
df = web.DataReader([ticker], 'yahoo', start, end)
df['Adj Close'].plot()
plt.show()
import datetime as dt
import matplotlib.pyplot as plt
from matplotlib import style
import pandas as pd
import pandas_datareader.data as web
style.use('ggplot')
start = dt.datetime(2000, 1, 1)
end = dt.datetime(2016, 12, 31)
df = web.DataReader('TSLA', 'yahoo', start, end)
print(df.tail(6))
def reset(self):
stock_code = np.random.choice(self.code_list)
s = pd.date_range(self.min_date, self.max_date, freq="D")
sample_idx = random.randrange(self.window_size, len(s) - self.period)
start_date = str(s[sample_idx - self.window_size]).split()[0]
end_date = str(s[sample_idx + self.period]).split()[0]
self.src_data = data.DataReader(stock_code, 'yahoo', start_date,
end_date)
#프리프로세싱
#data = source['Close']
sdata = self.src_data
#ret = data.pct_change(1)
log_ret = np.log(sdata / sdata.shift(1))
log_ret.columns = [
'log_h', 'log_l', 'log_o', 'log_c', 'log_v', 'log_adj'
]
def min_max_norm(wdata):
return (wdata[-1] - wdata.min()) / (wdata.max() - wdata.min())
def mean_std_norm(wdata):
return (wdata[-1] - wdata.mean()) / wdata.std()
def svd_whiten(X):
# a = source[X]
U, s, Vt = np.linalg.svd(X, full_matrices=False)
# U and Vt are the singular matrices, and s contains the singular values.
# Since the rows of both U and Vt are orthonormal vectors, then U * Vt
# will be white
X_white = np.dot(U, Vt)
return X_white[-1]
def rolling_whiten(src, window=20, min_periods=10):
ret = []
for i in range(src.__len__()):
if i < min_periods - 1:
ret.append([0 for i in range(src.columns.__len__())])
elif i < window:
ret.append(svd_whiten(src[0:i].values))
else:
ret.append(svd_whiten(src[i - window:i].values))
pdata = pd.DataFrame(np.stack(ret, 0))
pdata.columns = [
'w' + str(i) for i in range(src.columns.__len__())
]
pdata.index = src.index
return pdata
norm0 = sdata.rolling(window=self.window_size,
min_periods=10).apply(min_max_norm,
raw=True).fillna(0)
norm1 = sdata.rolling(window=self.window_size,
min_periods=10).apply(mean_std_norm,
raw=True).fillna(0)
norm2 = rolling_whiten(
self.src_data, window=self.window_size, min_periods=10).fillna(
0) #pca whitening using close open min max vol
norm0.columns = ['mm_h', 'mm_l', 'mm_o', 'mm_c', 'mm_v', 'mm_adj']
norm1.columns = ['ms_h', 'ms_l', 'ms_o', 'ms_c', 'ms_v', 'ms_adj']
self.prep_data = pd.concat(
[self.src_data, norm0, norm1, norm2, log_ret], axis=1)
self.prev_action = 0
self.count = self.window_size
self.balance = self.init_money
self.num_stocks = 0
self.sum_action = 0
# 학습 데이터 분리
# features_training_data = [
# 'open_lastclose_ratio', 'high_close_ratio', 'low_close_ratio',
# 'close_lastclose_ratio', 'volume_lastvolume_ratio',
# 'close_ma5_ratio', 'volume_ma5_ratio',
# 'close_ma10_ratio', 'volume_ma10_ratio',
# 'close_ma20_ratio', 'volume_ma20_ratio',
# 'close_ma60_ratio', 'volume_ma60_ratio',
# 'close_ma120_ratio', 'volume_ma120_ratio'
# ]
da0 = self.prep_data.iloc[self.count]
state = torch.from_numpy(da0.values).float()
# state = torch.cat([state, torch.Tensor([self.sum_action]).view(1,-1)],dim=1)
return state
Created on Mon Aug 6 21:39:41 2018
@author: kennedy
"""
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from datetime import datetime
import pandas_datareader.data as web
from sklearn.model_selection import KFold, GridSearchCV
#
start_date = datetime(2016, 1, 1)
end_date = datetime(2018, 7, 16)
data = web.DataReader('IBM', "yahoo", start_date, end_date)
#define the feature vector we would be using for
#to plot our regression
df = data[['Open']]
df['Volatility'] = df['Open'] - df['Open'].shift(1).fillna(0)
#SVM model
from sklearn.svm import SVR
#this we would be using to draw our regression line
Xf1 = np.arange(1, len(df) + 1)
#Xf2 = (Xf1**2).astype(np.float64)
#Xf3 = (Xf1**3).astype(np.float64)
#Xf4 = (Xf1**4).astype(np.float64)
from sklearn import tree, svm, linear_model
import datetime as dt
import matplotlib.pyplot as plt
from matplotlib import style
import numpy as np
import pandas as pd
import pandas_datareader.data as web
start = dt.datetime(2015, 1, 1)
end = dt.datetime(2016, 12, 31)
ticker = 'RBC'
# Loading Data
df = web.DataReader(ticker, 'iex', start, end)
df = df.reset_index()
closing_price = []
highest_price = []
lowest_price = []
opening_price = []
volume_traded = []
dates = []
labels = []
for i in df[['close']]:
for j in df[i]:
closing_price.append(round(j, 2))
for i in df[['high']]:
import pandas as pd
from pandas import Series, DataFrame
import numpy as np
import pandas_datareader.data as web
import matplotlib.pyplot as plt
import seaborn as sns
sns.set_style('whitegrid') # Give nice white background with grid
from datetime import datetime
stocks = ['AAPL', 'TEF', 'MSFT', 'TSLA']
end = datetime.now()
start = datetime(end.year - 1, end.month, end.day)
main_df = web.DataReader(stocks, "google", start, end)['Close']
rets = main_df.pct_change()
rets = rets.dropna()
corr = main_df.corr()
mask = np.zeros_like(corr, dtype=np.bool)
mask[np.triu_indices_from(mask)] = True
cmap = sns.diverging_palette(220, 10, as_cmap=True)
sns.heatmap(corr,
mask=mask,
cmap=cmap,
vmax=1,
center=0,
annot=True,
square=True,
linewidths=.5,
cbar_kws={"shrink": .5})
avg_first = 0
avg_end_val = 0
avg_realised_return = 0
avg_discrep_perc = 0
avg_expected_value = 0
print("\nAlgorithm\n\n")
for year in time_frame:
start_date = "{}/01/2015".format(year)
end_date = "{}/01/2019".format(year)
try:
data = web.DataReader(stocks,
data_source="yahoo",
start=start_date,
end=end_date)['Adj Close']
data = data.dropna()
data.reset_index(inplace=True, drop=False)
except:
print("Testing")
data['Total'] = 0
i = 0
for tick in stocks:
data['Total'] = data['Total'] + data[tick] * w[i]
i += 1
data['pct_change'] = data['Total'].pct_change()
""" Name : c8_05_print_obs_from_Google.py Book : Python for Finance (2nd ed.) Publisher: Packt Publishing Ltd. Author : Yuxing Yan Date : 6/6/2017 email : [email protected] [email protected] """ import pandas_datareader.data as web import datetime ticker='WMT' begdate = datetime.datetime(2010, 1, 1) enddate= datetime.datetime(2015, 5, 9) x=web.DataReader(ticker,'yahoo-actions',begdate,enddate) print(x.head())
import python.pandas
#import pandas_datareader as pdr
#import datetime
import matplotlib
from pandas_datareader import data, wb
from datetime import date
# aapl = pdr.get_data_yahoo('AAPL',
# start=datetime.datetime(2019, 1, 1),
# end=datetime.datetime(2019, 2, 19))
#
# print(aapl)
start = date(2019, 4, 1)
end = date(2019, 4, 25)
df = data.DataReader('GE', 'yahoo', start, end)
print(df.head())
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from pandas_datareader import data as wb
tickers = ['PG', 'BEI.DE']
sec_data = pd.DataFrame()
for t in tickers:
sec_data[t] = wb.DataReader(t, data_source='yahoo',
start='2007-1-1')['Adj Close']
#Vamos utilizar a taxa de retorno logaritmica
#O desvio padrao dos retornos de uma empresa tambem pode ser chamado de risco ou volatilidade
#Uma acao que mostra um grande desvio de sua media, é chamada de mais volatil
sec_returns = np.log(sec_data / sec_data.shift(1))
#Agora iremos calcular a media da taxa de retorno log.
sec_returns['PG'].mean()
#Agora iremos calcular a media da taxa de retorno log. anual
sec_returns['PG'].mean() * 250
#agora iremos calcular o desvio padrao, que é calculado com o comando .std()
sec_returns['PG'].std()
sec_returns['PG'].std() * 250**0.5
#Agora iremos calcular a media da taxa de retorno log.
sec_returns['BEI.DE'].mean()
import pandas_datareader.data as web
import datetime
import pandas_datareader
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from pandas.plotting import scatter_matrix
import mplfinance as mpf
from matplotlib.dates import DateFormatter, date2num, WeekdayLocator, DateLocator, MONDAY
start = datetime.datetime(2012, 1, 1)
end = datetime.datetime(2019, 2, 1)
tesla = web.DataReader('TSLA', 'yahoo', start, end)
ford = web.DataReader('FORD', 'yahoo', start, end)
gm = web.DataReader('GM', 'yahoo', start, end)
# print(tesla.head())
# print(ford.head())
# print(gm.head())
# Plot based on opening prices
# tesla['Open'].plot(label='Tesla', figsize=(16, 8), title='Open Price')
# ford['Open'].plot(label='FORD')
# gm['Open'].plot(label='GM')
# # plt.legend()
# plt.show()
# *************
# Plot based on closing prices
# tesla['Adj Close'].plot(label='Tesla', figsize=(16, 8),
import pandas as pd from pandas_datareader import data, wb import datetime import matplotlib.pyplot as plt from sklearn import linear_model #List with stocks to analyze. stocks =['BCBA:BMA'] start = datetime.date(2017,1,1) end = datetime.date(2017,10,6) #Read 'OLHC' data from google finance data = data.DataReader(stocks[0],'google',start,end) #Linear regression for close price. #Split train and test sets X = data.drop(['Close'],1) y = data['Close'] X_train = X[:-50] X_test = X[-50:] X_test.dropna(inplace=True) y_train = y[:-50] y_test = y[-50:] reg = linear_model.LinearRegression() reg.fit(X_train,y_train)
import datetime
import pandas_datareader.data as web
df_stockload = web.DataReader("600797.SS", "yahoo",
datetime.datetime(2018, 1, 1),
datetime.datetime(2019, 1, 1))
print(df_stockload.info())
# 替换 import matplotlib.finance as mpf 画k线图
import mpl_finance as mpf # 替换 import matplotlib.finance as mpf
import matplotlib.pyplot as plt
# 创建fig对象
fig = plt.figure(figsize=(8, 6), dpi=100, facecolor="white")
# 设置图像边框
fig.subplots_adjust(left=0.09,
bottom=0.20,
right=0.94,
top=0.90,
wspace=0.2,
hspace=0)
# 创建子图
graph_KAV = fig.add_subplot(1, 1, 1)
# 画k线
mpf.candlestick2_ochl(graph_KAV,
df_stockload.Open,
import matplotlib.pyplot as plt
import pandas_datareader.data as web
sk_hynix = web.DataReader("000660.KS", "yahoo")
fig = plt.figure(figsize=(12, 8))
top_axes = plt.subplot2grid((4, 4), (0, 0), rowspan=3, colspan=4)
bottom_axes = plt.subplot2grid((4, 4), (3, 0), rowspan=1, colspan=4)
bottom_axes.get_yaxis().get_major_formatter().set_scientific(False)
top_axes.plot(sk_hynix.index, sk_hynix['Adj Close'], label='Adjusted Close')
bottom_axes.plot(sk_hynix.index, sk_hynix['Volume'])
plt.tight_layout()
plt.show()
import pandas_datareader.data as pdr
import matplotlib.pyplot as plt
price = pdr.DataReader("^N225", 'yahoo', "1984/1/4", "2019/8/8")
#月の初めの値を選択
print(price.resample('M').first().tail())
#月の終わりの値を選択
print(price.resample('M').last().tail())
# 1日分オフセット
print(price.resample('M', loffset='1d').last().tail())
price.resample('A').Close().plot(color='magenta')
plt.ylabel('N225 Index')
plt.show()
def analyzepair(self, pairlist, startdate, enddate, showplot):
s1 = pairlist[0]
s2 = pairlist[1]
print '---------------------'
print 'Doing', s1, s2
#startdatetime = datetime.datetime(2015, 1, 1)
#enddatetime = datetime.datetime(2017, 9, 30)
startdatetime = datetime.datetime.strptime(startdate, "%Y-%m-%d")
enddatetime = datetime.datetime.strptime(enddate, "%Y-%m-%d")
#print startdatetime
#print enddatetime
#stop
prices1 = data.DataReader(s1, "yahoo", startdatetime, enddatetime)
prices2 = data.DataReader(s2, "yahoo", startdatetime, enddatetime)
df_a = pd.DataFrame(index=prices1.index)
df_a[s1] = prices1["Close"]
df_a[s2] = prices2["Close"]
df = df_a.dropna(axis=0, how='any')
#print df
if showplot == True:
# Plot the two time series
self.plot_price_series(df, s1, s2, startdatetime, enddatetime)
# Display a scatter plot of the two time series
self.plot_scatter_series(df, s1, s2)
Y = df[s2].tolist()
X = df[s1].tolist()
#print Y[:5]
#stop
#print 'got here 1'
# Calculate optimal hedge ratio "beta"
#print Y
#print X
#print 'got here 3'
#beta_hr = results
#print 'beta_hr'
#print beta_hr
# Calculate the residuals of the linear combination
i0 = 0
beta_hr_list = []
for idx, rows in df.iterrows():
#print idx
if i0 >= 0:
res = sm.OLS(Y[:i0 + 1], X[:i0 + 1])
beta_hr = res.fit().params[0]
mydict = {
'Date': idx,
'beta_hr': beta_hr,
'actual': Y[i0] / X[i0]
}
beta_hr_list.append(mydict)
i0 += 1
#if i0 >= 10:
# stop
df_beta_hr = pd.DataFrame(beta_hr_list)
df_beta_hr.set_index("Date", drop=True, inplace=True)
df_b = pd.concat([df_a, df_beta_hr], axis=1)
df_b['a-b'] = df_b['actual'] - df_b['beta_hr']
df_b["res"] = df_b[s2] - df_b['beta_hr'] * df_b[s1]
## for idx, row in df_b.iterrows():
## print idx,row['beta_hr'],round(row['a-b'],4)
# Calculate and analyze the CADF test on the residuals
cadf = ts.adfuller(df_b["res"])
test_null_hypothesis = cadf[0]
five_percent_value = cadf[4]['5%']
print ''
print 'cadf test for cointegration result 5% value:', test_null_hypothesis, 'must be less than', five_percent_value
return df_b
def test_iex_bad_symbol(self):
with pytest.raises(Exception):
web.DataReader("BADTICKER", "iex,", self.start, self.end)
import pandas as pd
import numpy as np
import datetime
import matplotlib.pyplot as plt
#%matplotlib inline
pd.set_option('display.notebook_repr_html', False)
pd.set_option('display.max_columns', 15)
pd.set_option('display.max_rows', 8)
pd.set_option('precision', 3)
from pandas_datareader import data as web
start = datetime.datetime(2012, 1, 1)
end = datetime.datetime(2012, 12, 30)
msft = web.DataReader("MSFT", "yahoo", start, end)
aapl = web.DataReader("AAPL", "yahoo", start, end)
def test_daily_invalid_date(self):
start = datetime(2000, 1, 5)
end = datetime(2017, 5, 24)
with pytest.raises(Exception):
web.DataReader(["AAPL", "TSLA"], "iex", start, end)
name: str = self["name"]
category: str = self["category"]
subcategories: dict = self["subcategories"]
metatype: str = self["metatype"]
submetatype: str = self["submetatype"]
abbreviation: str = self["abbreviation"]
jscat = self.main_helper.generate_hash(subcategories)
return f"{name}:{category}:{jscat}:{metatype}:{submetatype}:{abbreviation}"
if __name__ == "__main__":
import pandas_datareader.data as web
data_msft = web.DataReader(
"MSFT", "yahoo", start="2010/1/1", end="2020/1/30"
).round(2)
data_apple = web.DataReader(
"AAPL", "yahoo", start="2010/1/1", end="2020/1/30"
).round(2)
# print(data_apple)
episode_id = uuid.uuid4().hex
jambo = Jamboree()
data_hander = DataHandler()
data_hander.event = jambo
data_hander.processor = jambo
# The episode and live parameters are probably not good for the scenario. Will probably need to switch to something else to identify data
data_hander.episode = episode_id
data_hander.live = False
data_hander["category"] = "markets"
data_hander["subcategories"] = {
def test_multiple_symbols(self):
syms = ["AAPL", "MSFT", "TSLA"]
df = web.DataReader(syms, "iex", self.start, self.end)
assert sorted(list(df.columns.levels[1])) == syms
for sym in syms:
assert len(df.xs(sym, level="Symbols", axis=1) == 578)
def basic():
df = data.DataReader("PTT.BK", data_source="yahoo",
start="2017-1-1", end="2017-1-31")
print(df.head())
df.to_csv("ptt.csv")
