In [168]:

# Get FTSE100 index back from 01-Jan-1999
import html5lib
base_url = 'https://www.google.com/finance/historical?cid=12590587&startdate;=Jan%201%2C%201999&' + \
            'enddate=Oct%2015%2C%202016#=200&ei;=qVQCWMmRHMf7UKqMn7AM&start;='
ftse = pd.DataFrame()
for page in range(0,85,200):  #4485
    url = base_url + str(page)
    print('getting- ', url[:50] + '...' + url[110:])
    ftse = ftse.append(pd.read_html(url, flavor='html5lib', attrs={'class': 'gf-table historical_price'}, 
                                    parse_dates=[0], header=0)[0])
ftse.set_index('Date')

getting-  https://www.google.com/finance/historical?cid=1259...num=200&ei;=qVQCWMmRHMf7UKqMn7AM&start;=0

In [182]:

ftse.head()

Out[182]:

	Open	High	Low	Close	Volume
Date
2016-10-14	6977.740	7055.180	6977.740	7013.550	-
2016-10-13	7024.010	7024.010	6930.040	6977.740	-
2016-10-12	7070.880	7079.420	7015.760	7024.010	-
2016-10-11	7097.500	7129.830	7061.360	7070.880	-
2016-10-10	7044.390	7104.490	7024.080	7097.500	-

In [56]:

%matplotlib inline
import quandl
import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import numpy as np
np.set_printoptions(supress=True) 
pd.set_option('display.float_format', lambda x: '%.4f' % x)

barclays = quandl.get("GOOG/LON_BARC", authtoken="TMMe4w92TaVmnEgC1D6z")
barclays['Gain'] = barclays['Close'] - barclays['Open']
barclays['Day'] = barclays.index.dayofweek
barclays = barclays[(barclays.Volume!=0) & (barclays.Day!=5)]
barclays.head()

Out[56]:

	Open	High	Low	Close	Volume	Gain	Day
Date
1999-01-04	322.0	328.00	319.0	323.0	38524204.0	1.0	0
1999-01-05	322.5	338.25	320.5	326.0	22667444.0	3.5	1
1999-01-06	342.0	361.20	340.0	356.9	33591692.0	14.9	2
1999-01-07	354.1	357.20	325.0	351.9	65948960.0	-2.2	3
1999-01-08	359.0	369.00	334.0	368.0	28346880.0	9.0	4

In [ ]:

plt.figure(figsize=(20,4))
plt.subplot(1,2,1)
barclays.Close.plot()
ftse.Close.plot()

In [ ]:

In [4]:

sns.kdeplot(barclays['Close']);

C:\Users\Graham\Anaconda3\lib\site-packages\statsmodels\nonparametric\kdetools.py:20: VisibleDeprecationWarning: using a non-integer number instead of an integer will result in an error in the future
  y = X[:m/2+1] + np.r_[0,X[m/2+1:],0]*1j

In [5]:

sns.distplot(barclays['Close'])

C:\Users\Graham\Anaconda3\lib\site-packages\statsmodels\nonparametric\kdetools.py:20: VisibleDeprecationWarning: using a non-integer number instead of an integer will result in an error in the future
  y = X[:m/2+1] + np.r_[0,X[m/2+1:],0]*1j

Out[5]:

In [6]:

sns.lmplot('Open', 'Close', data=barclays, fit_reg=False)

Out[6]:

In [67]:

barclays_gr = barclays.groupby('Day')

In [75]:

barclays_gr.Gain.mean()

Out[75]:

Day
0   -0.508
1    0.054
2    0.126
3   -0.401
4   -0.316
Name: Gain, dtype: float64

In [109]:

plt.figure(figsize=(14,4))
plt.subplot(1,2,1)
barclays_gr.Gain.mean().plot(kind='bar')
plt.subplot(1,2,2)
barclays_gr.Volume.mean().plot(kind='bar')

Out[109]:

In [ ]:

In [116]:

import pandas_datareader.data as web
import datetime

start = datetime.datetime(2010, 1, 1)
end = datetime.datetime(2013, 1, 27)
f1 = web.DataReader("GOOG", 'yahoo', start, end)
#f2 = web.DataReader("INDEXFTSE:UKX", 'google', start, end)

In [1]:

%matplotlib inline
import numpy as np
import matplotlib.pyplot as plt
from scipy import signal
t = np.linspace(0, 5, 100)
x = t + np.random.normal(size=100)
xde = signal.detrend(x)
diff = x - xde

plt.plot(t, x, linewidth=1) 
plt.plot(t, xde, linewidth=1) 
plt.plot(t, diff, linewidth=1)

Out[1]:

[]

In [ ]:

%matplotlib notebook
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import matplotlib.animation as animation
import requests
import time
import datetime as dt
from IPython.display import display
from ipywidgets import widgets

def get_price(ticker):
    resp = requests.get(r"http://finance.yahoo.com/webservice/v1/symbols/{}/quote?format=json" \
                            .format(ticker), headers={'User-Agent': 'Mozilla/5.0'}).json()
    return float(resp['list']['resources'][0]['resource']['fields']['price'])

def animate(frame):
    global x, y
    x.append( dt.datetime.now() )
    price = get_price(ticker)
    y.append( price)
    ax.set_xlim([max(x) - dt.timedelta(seconds=65), max(x)])
    ax.set_ylim([min(y) * 0.99999, max(y) * 1.00001])
    line.set_data(x, y)
    message = '{} {}'.format(ticker, price)
    widget.value = '{} => {}
'.format(dt.datetime.now(), message)
    time.sleep(2)
    return line,

ticker = 'GBPUSD=X'
fig = plt.figure(num='Yahoo Finance Ticker: '+ticker)
ax = plt.axes()
ax.grid(True)
ax.get_yaxis().get_major_formatter().set_useOffset(False)
fig.autofmt_xdate()
time_locator = mdates.SecondLocator(interval=5)
time_fmt = mdates.DateFormatter('%H:%M:%S')
ax.xaxis.set_major_locator(time_locator)
ax.xaxis.set_major_formatter(time_fmt)
line, = ax.plot([], [], lw=1)

x=[]
y=[]
widget = widgets.HTML(value='')
display(widget)

anim = animation.FuncAnimation(fig, animate, repeat=False, frames=30, interval=50, blit=False)
plt.show()

GCHESTER.COM

Basic Stock price tinkering with iPython

{} => {}

Comments