In [120]:
import numpy as np
import pandas as pd
import Quandl
from matplotlib import pyplot as plt
from sklearn.linear_model import BayesianRidge
%matplotlib inline
In [121]:
# http://www.quantshare.com/sa-426-6-ways-to-download-free-intraday-and-tick-data-for-the-us-stock-market
In [122]:
# http://www.google.com/finance/getprices?i=300&p=10d&f=d,o,h,l,c,v&df=cpct&q=TNXP
In [123]:
!head tnxp_goog_5min_10day_clean.csv
In [124]:
tnxp_df = pd.read_csv("tnxp_goog_5min_10day_clean.csv")
In [125]:
print tnxp_df.shape
tnxp_df.head()
Out[125]:
In [126]:
tnxp_df.describe()
Out[126]:
In [127]:
tnxp_df['CLOSE'].plot()
plt.show()
In [128]:
close_values = tnxp_df['LOW'].values[:-1]
low_values = tnxp_df['LOW'].values[:-1]
high_values = tnxp_df['HIGH'].values[:-1]
vol_values = tnxp_df['VOLUME'].values[:-1]
In [129]:
tnxp_df['LASTCLOSE'] = np.insert(close_values, 0, 0.0)
tnxp_df['LASTLOW'] = np.insert(low_values, 0, 0.0)
tnxp_df['LASTHIGH'] = np.insert(high_values, 0, 0.0)
tnxp_df['LASTVOLUME'] = np.insert(vol_values, 0, 0.0)
In [130]:
tnxp_df = tnxp_df.ix[1:]
In [131]:
tnxp_df.head()
Out[131]:
In [132]:
tnxp_df.tail()
Out[132]:
In [133]:
tnxp_df['LOWPRICEIN30MINS'] = np.append(tnxp_df['LOW'].values[6:], [0,0,0,0,0,0])
In [134]:
tnxp_df.head(10)
Out[134]:
In [135]:
tnxp_df.tail(10)
Out[135]:
In [150]:
print [str(col) for col in train_df.columns]
In [151]:
train_df = tnxp_df.ix[0:768]
In [153]:
X = train_df[['INDEX', 'CLOSE', 'HIGH', 'LOW', 'OPEN', 'VOLUME', 'LASTCLOSE', 'LASTLOW', 'LASTHIGH', 'LASTVOLUME']].values
y = train_df[['LOWPRICEIN30MINS']].values.flatten()
print X.shape, y.shape
In [154]:
clf = BayesianRidge(compute_score=True)
clf.fit(X, y)
Out[154]:
In [183]:
value_index = 1
features = tnxp_df.ix[value_index].values
features = np.delete(features, features.shape[0]-1)
prediction = clf.predict(features)
print "Actual: {}".format(tnxp_df.ix[value_index].values[-1])
print "Prediction: {}".format(prediction[0])
print "Difference: {}".format(tnxp_df.ix[value_index].values[-1] - prediction[0])
In [186]:
value_index = 768
features = tnxp_df.ix[value_index].values
features = np.delete(features, features.shape[0]-1)
prediction = clf.predict(features)
print "Actual: {}".format(tnxp_df.ix[value_index].values[-1])
print "Prediction: {}".format(prediction[0])
print "Difference: {}".format(tnxp_df.ix[value_index].values[-1] - prediction[0])
In [188]:
value_index = 769
features = tnxp_df.ix[value_index].values
features = np.delete(features, features.shape[0]-1)
prediction = clf.predict(features)
print "Actual: {}".format(10.40)
print "Prediction: {}".format(prediction[0])
print "Difference: {}".format(10.40 - prediction[0])
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [ ]:
In [123]:
# june 17
prediction = clf.predict(np.array([9.97, 9.67, 797492, 9.90]))
print prediction, prediction - (prediction*0.05)
In [125]:
# june 17
prediction = clf.predict(np.array([9.97, 9.67, 797400, 9.90]))
print prediction, prediction - (prediction*0.05)
In [ ]:
In [ ]:
In [ ]:
In [72]:
X = tnxp_df[['Open', 'High', 'Low', 'Volume']].values
y = tnxp_df[['Close']].values.flatten()
print X.shape, y.shape
In [73]:
indices = np.random.permutation(np.arange(0,X.shape[0]-1))
In [74]:
train_count = np.floor(X.shape[0] * 0.80)
print train_count
print np.ceil(X.shape[0] * 0.20)
In [75]:
X_train = X[:train_count]
y_train = y[:train_count]
X_test = X[train_count:]
y_test = y[train_count:]
print len(X_train), len(X_test), len(y_train), len(y_test)
In [76]:
clf = BayesianRidge(compute_score=True)
clf.fit(X_train, y_train)
Out[76]:
In [77]:
clf.score(X_test, y_test)
Out[77]:
In [78]:
clf.predict(np.array([9.97, 10.45, 9.67, 797492]))
Out[78]:
In [79]:
clf.predict(np.array([7.18, 7.18, 6.79, 123700]))
Out[79]:
In [ ]:
In [8]:
print(__doc__)
import numpy as np
import matplotlib.pyplot as plt
from scipy import stats
from sklearn.linear_model import BayesianRidge, LinearRegression
###############################################################################
# Generating simulated data with Gaussian weigthts
np.random.seed(0)
n_samples, n_features = 100, 100
X = np.random.randn(n_samples, n_features) # Create Gaussian data
# Create weigts with a precision lambda_ of 4.
lambda_ = 4.
w = np.zeros(n_features)
# Only keep 10 weights of interest
relevant_features = np.random.randint(0, n_features, 10)
for i in relevant_features:
w[i] = stats.norm.rvs(loc=0, scale=1. / np.sqrt(lambda_))
# Create noise with a precision alpha of 50.
alpha_ = 50.
noise = stats.norm.rvs(loc=0, scale=1. / np.sqrt(alpha_), size=n_samples)
# Create the target
y = np.dot(X, w) + noise
###############################################################################
# Fit the Bayesian Ridge Regression and an OLS for comparison
clf = BayesianRidge(compute_score=True)
clf.fit(X, y)
ols = LinearRegression()
ols.fit(X, y)
###############################################################################
# Plot true weights, estimated weights and histogram of the weights
plt.figure(figsize=(6, 5))
plt.title("Weights of the model")
plt.plot(clf.coef_, 'b-', label="Bayesian Ridge estimate")
plt.plot(w, 'g-', label="Ground truth")
plt.plot(ols.coef_, 'r--', label="OLS estimate")
plt.xlabel("Features")
plt.ylabel("Values of the weights")
plt.legend(loc="best", prop=dict(size=12))
plt.figure(figsize=(6, 5))
plt.title("Histogram of the weights")
plt.hist(clf.coef_, bins=n_features, log=True)
plt.plot(clf.coef_[relevant_features], 5 * np.ones(len(relevant_features)),
'ro', label="Relevant features")
plt.ylabel("Features")
plt.xlabel("Values of the weights")
plt.legend(loc="lower left")
plt.figure(figsize=(6, 5))
plt.title("Marginal log-likelihood")
plt.plot(clf.scores_)
plt.ylabel("Score")
plt.xlabel("Iterations")
plt.show()
In [4]:
#https://www.google.com/finance/getprices?i=1&p=1000d&f=d,o,h,l,c,v&df=cpct&q=LPTH
In [2]:
tickers_df = pd.read_csv('WIKI_tickers.csv')
tickers_df.shape
Out[2]:
In [3]:
tickers = tickers_df['quandl code'].values.tolist()
print len(tickers)
print tickers[:5]
In [6]:
new_tickers = []
for ticker in tickers[5:10]:
stock = Quandl.get(ticker, authtoken='DVhizWXNTePyzzy1eHWR')
if stock['Close'].values[-1] < 20.00:
new_tickers.append(ticker)
print len(new_tickers)
In [7]:
new_tickers
Out[7]:
In [8]:
stock = Quandl.get(new_tickers[0], authtoken='DVhizWXNTePyzzy1eHWR')
In [9]:
stock.tail()
Out[9]:
In [54]:
plt.subplot(2,1,1)
plt.plot(stock.index[-10:],stock['Close'].values[-10:])
plt.subplot(2,1,2)
plt.bar(stock.index[-10:],stock['Volume'].values[-10:])
plt.xticks(rotation=45)
plt.show()
In [29]:
plt.plot(stock.index[-10:],stock['Close'].values[-10:])
plt.show()
In [30]:
plt.bar(stock.index[-10:], stock['Volume'].values[-10:])
Out[30]:
In [63]:
#!/usr/bin/env python
import matplotlib.pyplot as plt
from matplotlib.dates import DateFormatter, WeekdayLocator, DayLocator, MONDAY
from matplotlib.finance import quotes_historical_yahoo_ohlc, candlestick_ohlc
# (Year, month, day) tuples suffice as args for quotes_historical_yahoo
date1 = (2015, 5, 15)
date2 = (2015, 6, 15)
mondays = WeekdayLocator(MONDAY) # major ticks on the mondays
alldays = DayLocator() # minor ticks on the days
weekFormatter = DateFormatter('%b %d') # e.g., Jan 12
dayFormatter = DateFormatter('%d') # e.g., 12
quotes = quotes_historical_yahoo_ohlc('AES', date1, date2)
if len(quotes) == 0:
raise SystemExit
fig, ax = plt.subplots()
fig.subplots_adjust(bottom=0.2)
ax.xaxis.set_major_locator(mondays)
ax.xaxis.set_minor_locator(alldays)
ax.xaxis.set_major_formatter(weekFormatter)
#ax.xaxis.set_minor_formatter(dayFormatter)
#plot_day_summary(ax, quotes, ticksize=3)
candlestick_ohlc(ax, quotes, width=0.6, colorup='g', colordown='r')
ax.xaxis_date()
ax.autoscale_view()
plt.setp(plt.gca().get_xticklabels(), rotation=45, horizontalalignment='right')
plt.show()
In [64]:
import datetime
import numpy as np
import matplotlib.colors as colors
import matplotlib.finance as finance
import matplotlib.dates as mdates
import matplotlib.ticker as mticker
import matplotlib.mlab as mlab
import matplotlib.pyplot as plt
import matplotlib.font_manager as font_manager
startdate = datetime.date(2006,1,1)
today = enddate = datetime.date.today()
ticker = 'SPY'
fh = finance.fetch_historical_yahoo(ticker, startdate, enddate)
# a numpy record array with fields: date, open, high, low, close, volume, adj_close)
r = mlab.csv2rec(fh); fh.close()
r.sort()
def moving_average(x, n, type='simple'):
"""
compute an n period moving average.
type is 'simple' | 'exponential'
"""
x = np.asarray(x)
if type=='simple':
weights = np.ones(n)
else:
weights = np.exp(np.linspace(-1., 0., n))
weights /= weights.sum()
a = np.convolve(x, weights, mode='full')[:len(x)]
a[:n] = a[n]
return a
def relative_strength(prices, n=14):
"""
compute the n period relative strength indicator
http://stockcharts.com/school/doku.php?id=chart_school:glossary_r#relativestrengthindex
http://www.investopedia.com/terms/r/rsi.asp
"""
deltas = np.diff(prices)
seed = deltas[:n+1]
up = seed[seed>=0].sum()/n
down = -seed[seed<0].sum()/n
rs = up/down
rsi = np.zeros_like(prices)
rsi[:n] = 100. - 100./(1.+rs)
for i in range(n, len(prices)):
delta = deltas[i-1] # cause the diff is 1 shorter
if delta>0:
upval = delta
downval = 0.
else:
upval = 0.
downval = -delta
up = (up*(n-1) + upval)/n
down = (down*(n-1) + downval)/n
rs = up/down
rsi[i] = 100. - 100./(1.+rs)
return rsi
def moving_average_convergence(x, nslow=26, nfast=12):
"""
compute the MACD (Moving Average Convergence/Divergence) using a fast and slow exponential moving avg'
return value is emaslow, emafast, macd which are len(x) arrays
"""
emaslow = moving_average(x, nslow, type='exponential')
emafast = moving_average(x, nfast, type='exponential')
return emaslow, emafast, emafast - emaslow
plt.rc('axes', grid=True)
plt.rc('grid', color='0.75', linestyle='-', linewidth=0.5)
textsize = 9
left, width = 0.1, 0.8
rect1 = [left, 0.7, width, 0.2]
rect2 = [left, 0.3, width, 0.4]
rect3 = [left, 0.1, width, 0.2]
fig = plt.figure(facecolor='white')
axescolor = '#f6f6f6' # the axes background color
ax1 = fig.add_axes(rect1, axisbg=axescolor) #left, bottom, width, height
ax2 = fig.add_axes(rect2, axisbg=axescolor, sharex=ax1)
ax2t = ax2.twinx()
ax3 = fig.add_axes(rect3, axisbg=axescolor, sharex=ax1)
### plot the relative strength indicator
prices = r.adj_close
rsi = relative_strength(prices)
fillcolor = 'darkgoldenrod'
ax1.plot(r.date, rsi, color=fillcolor)
ax1.axhline(70, color=fillcolor)
ax1.axhline(30, color=fillcolor)
ax1.fill_between(r.date, rsi, 70, where=(rsi>=70), facecolor=fillcolor, edgecolor=fillcolor)
ax1.fill_between(r.date, rsi, 30, where=(rsi<=30), facecolor=fillcolor, edgecolor=fillcolor)
ax1.text(0.6, 0.9, '>70 = overbought', va='top', transform=ax1.transAxes, fontsize=textsize)
ax1.text(0.6, 0.1, '<30 = oversold', transform=ax1.transAxes, fontsize=textsize)
ax1.set_ylim(0, 100)
ax1.set_yticks([30,70])
ax1.text(0.025, 0.95, 'RSI (14)', va='top', transform=ax1.transAxes, fontsize=textsize)
ax1.set_title('%s daily'%ticker)
### plot the price and volume data
dx = r.adj_close - r.close
low = r.low + dx
high = r.high + dx
deltas = np.zeros_like(prices)
deltas[1:] = np.diff(prices)
up = deltas>0
ax2.vlines(r.date[up], low[up], high[up], color='black', label='_nolegend_')
ax2.vlines(r.date[~up], low[~up], high[~up], color='black', label='_nolegend_')
ma20 = moving_average(prices, 20, type='simple')
ma200 = moving_average(prices, 200, type='simple')
linema20, = ax2.plot(r.date, ma20, color='blue', lw=2, label='MA (20)')
linema200, = ax2.plot(r.date, ma200, color='red', lw=2, label='MA (200)')
last = r[-1]
s = '%s O:%1.2f H:%1.2f L:%1.2f C:%1.2f, V:%1.1fM Chg:%+1.2f' % (
today.strftime('%d-%b-%Y'),
last.open, last.high,
last.low, last.close,
last.volume*1e-6,
last.close-last.open )
t4 = ax2.text(0.3, 0.9, s, transform=ax2.transAxes, fontsize=textsize)
props = font_manager.FontProperties(size=10)
leg = ax2.legend(loc='center left', shadow=True, fancybox=True, prop=props)
leg.get_frame().set_alpha(0.5)
volume = (r.close*r.volume)/1e6 # dollar volume in millions
vmax = volume.max()
poly = ax2t.fill_between(r.date, volume, 0, label='Volume', facecolor=fillcolor, edgecolor=fillcolor)
ax2t.set_ylim(0, 5*vmax)
ax2t.set_yticks([])
### compute the MACD indicator
fillcolor = 'darkslategrey'
nslow = 26
nfast = 12
nema = 9
emaslow, emafast, macd = moving_average_convergence(prices, nslow=nslow, nfast=nfast)
ema9 = moving_average(macd, nema, type='exponential')
ax3.plot(r.date, macd, color='black', lw=2)
ax3.plot(r.date, ema9, color='blue', lw=1)
ax3.fill_between(r.date, macd-ema9, 0, alpha=0.5, facecolor=fillcolor, edgecolor=fillcolor)
ax3.text(0.025, 0.95, 'MACD (%d, %d, %d)'%(nfast, nslow, nema), va='top',
transform=ax3.transAxes, fontsize=textsize)
#ax3.set_yticks([])
# turn off upper axis tick labels, rotate the lower ones, etc
for ax in ax1, ax2, ax2t, ax3:
if ax!=ax3:
for label in ax.get_xticklabels():
label.set_visible(False)
else:
for label in ax.get_xticklabels():
label.set_rotation(30)
label.set_horizontalalignment('right')
ax.fmt_xdata = mdates.DateFormatter('%Y-%m-%d')
class MyLocator(mticker.MaxNLocator):
def __init__(self, *args, **kwargs):
mticker.MaxNLocator.__init__(self, *args, **kwargs)
def __call__(self, *args, **kwargs):
return mticker.MaxNLocator.__call__(self, *args, **kwargs)
# at most 5 ticks, pruning the upper and lower so they don't overlap
# with other ticks
#ax2.yaxis.set_major_locator(mticker.MaxNLocator(5, prune='both'))
#ax3.yaxis.set_major_locator(mticker.MaxNLocator(5, prune='both'))
ax2.yaxis.set_major_locator(MyLocator(5, prune='both'))
ax3.yaxis.set_major_locator(MyLocator(5, prune='both'))
plt.show()
In [66]:
import numpy as np
import matplotlib
import matplotlib.pyplot as plt
from matplotlib.finance import candlestick
from matplotlib.finance import volume_overlay3
from matplotlib.dates import num2date
from matplotlib.dates import date2num
import matplotlib.mlab as mlab
import datetime
datafile = 'data.csv'
r = mlab.csv2rec(datafile, delimiter=';')
# the dates in my example file-set are very sparse (and annoying) change the dates to be sequential
for i in range(len(r)-1):
r['date'][i+1] = r['date'][i] + datetime.timedelta(days=1)
candlesticks = zip(date2num(r['date']),r['open'],r['close'],r['max'],r['min'],r['volume'])
fig = plt.figure()
ax = fig.add_subplot(1,1,1)
ax.set_ylabel('Quote ($)', size=20)
candlestick(ax, candlesticks,width=1,colorup='g', colordown='r')
# shift y-limits of the candlestick plot so that there is space at the bottom for the volume bar chart
pad = 0.25
yl = ax.get_ylim()
ax.set_ylim(yl[0]-(yl[1]-yl[0])*pad,yl[1])
# create the second axis for the volume bar-plot
ax2 = ax.twinx()
# set the position of ax2 so that it is short (y2=0.32) but otherwise the same size as ax
ax2.set_position(matplotlib.transforms.Bbox([[0.125,0.1],[0.9,0.32]]))
# get data from candlesticks for a bar plot
dates = [x[0] for x in candlesticks]
dates = np.asarray(dates)
volume = [x[5] for x in candlesticks]
volume = np.asarray(volume)
# make bar plots and color differently depending on up/down for the day
pos = r['open']-r['close']<0
neg = r['open']-r['close']>0
ax2.bar(dates[pos],volume[pos],color='green',width=1,align='center')
ax2.bar(dates[neg],volume[neg],color='red',width=1,align='center')
#scale the x-axis tight
ax2.set_xlim(min(dates),max(dates))
# the y-ticks for the bar were too dense, keep only every third one
yticks = ax2.get_yticks()
ax2.set_yticks(yticks[::3])
ax2.yaxis.set_label_position("right")
ax2.set_ylabel('Volume', size=20)
# format the x-ticks with a human-readable date.
xt = ax.get_xticks()
new_xticks = [datetime.date.isoformat(num2date(d)) for d in xt]
ax.set_xticklabels(new_xticks,rotation=45, horizontalalignment='right')
plt.ion()
plt.show()