def predict(request, symbol): """ Executed when a request for /predict/SYMBOL comes in. Django handles parsing this string for SYMBOL. The default time range is 2000 to now. Parameters ---------- request: incoming request symbol: stock symbol Returns ------- projection: data needed for "projection" div in index.html to show a next-day stock price via ajax """ # find the next day stock price in this time series start = datetime.date(2009, 12, 31) end = datetime.datetime.now() # get the stock history for that symbol with start and end dates stock = Stocks(symbol, start, end) dates, prices = stock.get_stock_history() # use our neural network to predict the next day closing price predictor = Predictors(dates, prices) projection = predictor.predict() projection = "%.2f" % projection return HttpResponse(projection, mimetype="text/html")
def index(request):
#csrf_token = get_token(request)
"""
Executed when a request for / comes in. The default symbol is 'EBAY'.
The default time range is 2000 to now.
Parameters
----------
request: incoming request
Returns
-------
context: data needed for template index.html to render
"""
symbol = ""
if request.method == "POST":
symbol = request.POST.get("symbol")
symbol = symbol.upper()
if symbol == "":
symbol = 'GOOG'
start = datetime.date(1999, 12, 31)
end = datetime.datetime.now()
# get the stock history for that symbol with start and end dates
stock = Stocks(symbol, start, end)
try:
dates, prices = stock.get_stock_history()
except Exception as e:
symbol = 'EBAY'
stock = Stocks(symbol, start, end)
dates, prices = stock.get_stock_history()
# find the MA and MACD of this time series
indicators = Indicators(prices)
ma20 = indicators.moving_average(7, type='simple')
macd = indicators.moving_average_convergence()
rsi = indicators.relative_strength()
template = 'datavis/index.html'
context = {
'symbol': symbol,
'dates': dates,
'prices': prices,
'ma20': ma20,
'macd': macd,
'rsi': rsi,
#'csrf_token': csrf_token
}
from django.template.context import RequestContext
return render_to_response(template, context,context_instance=RequestContext(request))
def kalman(request, symbol):
"""
Executed when a request for /kalman/SYMBOL comes in.
Django handles parsing this string for SYMBOL.
The default time range is 2000 to now.
Parameters
----------
request: incoming request
symbol: stock symbol
Returns
-------
context: data needed for template kalman.html to render
"""
# get the stock history for that symbol with start and end dates
start = datetime.date(1999, 12, 31)
end = datetime.datetime.now()
stock = Stocks(symbol, start, end)
dates, prices = stock.get_stock_history()
#get the mean and standard deviation for this price feature
mu = numpy.average(prices)
sigma = numpy.std(prices)
mu = "%.2f" % mu
sigma = "%.2f" % sigma
# apply the kalman filter to the time series
filters = Filters(prices, mu, sigma)
kalman = filters.kalmanfilter()
template = 'datavis/kalman.html'
context = {
'symbol': symbol,
'dates': dates,
'prices': prices,
'kalman': kalman,
'mu': mu,
'sigma': sigma,
}
return direct_to_template(request, template, context)
def fft(request, symbol):
"""
Executed when a request for /fft/SYMBOL comes in.
Django handles parsing this string for SYMBOL.
The default time range is 2000 to now.
Parameters
----------
request: incoming request
symbol: stock symbol
Returns
-------
context: data needed for template fft.html to render
"""
# get the stock history for that symbol with start and end dates
start = datetime.date(1999, 12, 31)
end = datetime.datetime.now()
stock = Stocks(symbol, start, end)
dates, prices = stock.get_stock_history()
# transform the time-series to the frequency domain
transform = Transforms(prices)
period, power = transform.fft2()
# reverse the numpy array
period = period[::-1]
power = power[::-1]
template = 'datavis/fft.html'
context = {
'symbol': symbol,
'period': period,
'power': power,
}
return render_to_response(template, context,context_instance=RequestContext(request))
