def svmr(stock_sym):
predicted_y_high = analyzer.analyzeSymbol(stock_sym)
long_svm_int = get_svm_prediction(stock_sym, predicted_y_high)
if (long_svm_int == +1):
long_svm = "strong"
else:
long_svm = "weak"
# Get 50 day MACD and compare with ANN prediction
long_macd = MACD.get_macd(stock_sym)
if (long_macd < predicted_y_high):
long_macd = "rising"
print("BUY")
else:
long_macd = "falling"
print("SELL")
return predicted_y_high, long_macd, long_svm
def training(variable):
key1 = analyzeSymbol(variable)
#key1 = dbtojsonfunction("abcd") # Use this to print graph in abcd.html inorder to print predict price and actual price
#key1 = dbtojsonfunction("abcd") # Use this to print graph in bank.html
#key1 = dbtojsonfunction("sanimabank") # Use this For stockanalysis.html
#
con = sql.connect("EBL.db")
cur = con.cursor()
cur.execute("SELECT * FROM {} ".format(variable, ))
individualdata = cur.fetchall()
#name = cur.fetchall()
#return render_template(os.path.join(path+'.html'), key1=key1, key2 = key2)
return render_template('training.html',
variable=variable,
individualdata=individualdata)
def apiAnalyze(stockSymbol):
return json.dumps(analyzeSymbol(stockSymbol))
comp_ticker = {"google": "GOOG", "apple": "AAPL"}
while True:
for mention in mentions:
tw = mention.text
_, cmd, comp = tw.split(' ')
print tw
print cmd, comp
print mention.user.screen_name
comp = comp.lower()
ticker = comp_ticker.get(comp, None)
price, coef = analyzer.analyzeSymbol(ticker)
market_fear = finance_scrape.get_fear_score(ticker)
print (price, coef, market_fear)
api.update_status(
"Hello @{}! I think that the {} stock will be at {:.2f} tomorrow - With a Confidence of {:.2f}% and a Bear ratio of {}".format(
mention.user.screen_name, comp.title(), price, 100 * coef, market_fear))
exit(0)
def getCompanymmand(phrase):
temp = phrase.split(' ')
command = temp[0].lower()
def apiAnalyze(stockSymbol):
return json.dumps(analyzeSymbol(stockSymbol))
def myprofile():
if request.cookies.get('login'):
username = str(request.cookies.get('login').split('_')[0])
else:
return redirect('http://127.0.0.1:5000/client/login')
cursor.execute("SELECT favorite_stock FROM login_info WHERE user_name = '{username}'".format(username=username))
focused_stock = str(cursor.fetchone()[0]).encode("utf-8")
cnx.commit()
try:
data_types = [
'High',
'Low',
'Close',
'Volume',
'Adj Close',
]
stockname = focused_stock
# stockname = request.args.get('chart_title')
window = 50
window2 = 150
recent_trend_query = """
SELECT Date, `{__value_name__}`
FROM seproject.historicaldata
where sym = '{__stockname__}'
order by Date ASC;
"""
move_avg_query = """
SELECT seproject.historicaldata.Date, seproject.historicaldata.`{__value_name__}`, avg(historicaldata_past.`{__value_name__}`) as `{__value_name__}_window`
FROM seproject.historicaldata
JOIN (
SELECT
seproject.historicaldata.Date, seproject.historicaldata.`{__value_name__}`
FROM seproject.historicaldata
WHERE seproject.historicaldata.sym = '{__stockname__}'
) AS historicaldata_past
ON seproject.historicaldata.Date BETWEEN historicaldata_past.Date and date_add(historicaldata_past.Date, interval + {__window__} day)
WHERE seproject.historicaldata.sym = '{__stockname__}'
GROUP BY 1, 2
order by seproject.historicaldata.Date ASC;
"""
rt_price = """
SELECT price
FROM seproject.realtimedata
WHERE sym = '{__stockname__}'
"""
stock_symb = [
'AAPL',
'GOOGL',
'NVDA',
'YHOO',
'AMZN',
'MSFT',
'BAC',
'NKE',
'NFLX',
'FB',
]
pred_price = round(analyzeSymbol(stockname, 5), 2)
pred_price1 = round(analyzeSymbol(stockname, 50), 2)
RSI = tuple(getRSI(stockname))
q_results = db.engine.execute(rt_price.format(__stockname__=stockname))
# d = json.dumps(q_results)
for res in q_results:
chart_data = res[0]
rt_price1 = chart_data
rec_BS_A = ['BUY', 'SELL', 'HOLD']
if float(rt_price1) * (0.99) > pred_price:
rec_BS = rec_BS_A[1]
elif float(rt_price1) * (1.01) < pred_price:
rec_BS = rec_BS_A[0]
else:
rec_BS = rec_BS_A[2]
if float(rt_price1) * (0.99) > pred_price1:
rec_BS1 = rec_BS_A[1]
elif float(rt_price1) * (1.01) < pred_price1:
rec_BS1 = rec_BS_A[0]
else:
rec_BS1 = rec_BS_A[2]
rss_url = "//rss.bloople.net/?url=https%3A%2F%2Fwww.google.com%2Ffinance%2Fcompany_news%3Fq%3DNASDAQ%3A" + stockname + "%26ei%3DXt77WInNHZLteNzahsAC%26output%3Drss&detail=-1&showtitle=false&type=js"
chart_data_all = {}
chart_data_all_sec = {}
chart_data_all_th = {}
chart_data_all_fth = {}
for data_type in data_types:
q_results = db.engine.execute(recent_trend_query.format(__stockname__=stockname, __value_name__=data_type))
chart_values = []
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[1]])
chart_data_all[data_type] = chart_values
q_results = db.engine.execute(
move_avg_query.format(__stockname__=stockname, __value_name__=data_type, __window__=window))
chart_values = []
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[2]])
chart_data_all_sec[data_type] = chart_values
q_results = db.engine.execute(
move_avg_query.format(__stockname__=stockname, __value_name__=data_type, __window__=window2))
chart_values = []
chart_values1 = []
i = 0;
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[2]])
chart_values1.append([unixtime, RSI[i]])
i = i + 1
chart_data_all_th[data_type] = chart_values
chart_data_all_fth[data_type] = chart_values1
return render_template("myprofile.html",stock=focused_stock, chart_title=stockname,
chart_data_all=chart_data_all, chart_data_all_sec=chart_data_all_sec,
chart_data_all_th=chart_data_all_th, data_types=data_types,
window=window, window2=window2, rt_price=rt_price1,
pred_price=pred_price, rec_BS=rec_BS, pred_price1=pred_price1, rec_BS1=rec_BS1,
stock_symb=stock_symb, rss_url=rss_url, chart_data_all_fth=chart_data_all_fth)
except IndexError:
return render_template("myprofile.html")
def stock():
data_types = [
'High',
'Low',
'Close',
'Volume',
'Adj Close',
]
stockname = request.args.get('stock_name')
#stockname = request.args.get('chart_title')
window = 50
window2 = 150
recent_trend_query = """
SELECT Date, `{__value_name__}`
FROM seproject.historicaldata
where sym = '{__stockname__}'
order by Date ASC;
"""
move_avg_query = """
SELECT seproject.historicaldata.Date, seproject.historicaldata.`{__value_name__}`, avg(historicaldata_past.`{__value_name__}`) as `{__value_name__}_window`
FROM seproject.historicaldata
JOIN (
SELECT
seproject.historicaldata.Date, seproject.historicaldata.`{__value_name__}`
FROM seproject.historicaldata
WHERE seproject.historicaldata.sym = '{__stockname__}'
) AS historicaldata_past
ON seproject.historicaldata.Date BETWEEN historicaldata_past.Date and date_add(historicaldata_past.Date, interval + {__window__} day)
WHERE seproject.historicaldata.sym = '{__stockname__}'
GROUP BY 1, 2
order by seproject.historicaldata.Date ASC;
"""
rt_price = """
SELECT price
FROM seproject.realtimedata
WHERE sym = '{__stockname__}'
"""
stock_symb = [
'AAPL',
'GOOGL',
'NVDA',
'YHOO',
'AMZN',
'MSFT',
'BAC',
'NKE',
'NFLX',
'SNAP',
]
pred_price = round(analyzeSymbol(stockname, 5),2)
pred_price1 = round(analyzeSymbol(stockname, 50),2)
# get Bayesian Prediction
filename = stockname + '_histdata.csv'
returnBayesianPrice = getBayesianCurveFit('./', filename, '3', '0')
# get SVM prediction
returnSVM = SVMpredict(stockname) # get Bayesian Prediction
RSI = tuple(getRSI(stockname))
q_results = db.engine.execute(rt_price.format(__stockname__=stockname))
#d = json.dumps(q_results)
for res in q_results:
chart_data = res[0]
rt_price1 = chart_data
rec_BS_A = ['BUY','SELL','HOLD']
if float(rt_price1)*(0.99) > pred_price:
rec_BS = rec_BS_A[1]
elif float(rt_price1)*(1.01) < pred_price:
rec_BS = rec_BS_A[0]
else:
rec_BS = rec_BS_A[2]
if float(rt_price1)*(0.99) > pred_price1:
rec_BS1 = rec_BS_A[1]
elif float(rt_price1)*(1.01) < pred_price1:
rec_BS1 = rec_BS_A[0]
else:
rec_BS1 = rec_BS_A[2]
rss_url = "//rss.bloople.net/?url=https%3A%2F%2Fwww.google.com%2Ffinance%2Fcompany_news%3Fq%3DNASDAQ%3A"+stockname+"%26ei%3DXt77WInNHZLteNzahsAC%26output%3Drss&detail=-1&showtitle=false&type=js"
chart_data_all = {}
chart_data_all_sec = {}
chart_data_all_th = {}
chart_data_all_fth = {}
for data_type in data_types:
q_results = db.engine.execute(recent_trend_query.format(__stockname__=stockname, __value_name__=data_type))
chart_values = []
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[1]])
chart_data_all[data_type] = chart_values
q_results = db.engine.execute(move_avg_query.format(__stockname__=stockname, __value_name__=data_type, __window__=window))
chart_values = []
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[2]])
chart_data_all_sec[data_type] = chart_values
q_results = db.engine.execute(move_avg_query.format(__stockname__=stockname, __value_name__=data_type, __window__=window2))
chart_values = []
chart_values1 = []
i = 0;
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[2]])
chart_values1.append([unixtime, RSI[i]])
i = i + 1
chart_data_all_th[data_type] = chart_values
chart_data_all_fth[data_type] = chart_values1
if request.cookies.get('login'):
return render_template("stockselect.html", display='true',chart_title=stockname,
chart_data_all=chart_data_all, chart_data_all_sec=chart_data_all_sec,
chart_data_all_th=chart_data_all_th, data_types=data_types,
window=window, window2= window2, rt_price= rt_price1,
pred_price = pred_price, rec_BS = rec_BS, pred_price1 = pred_price1, rec_BS1 = rec_BS1, stock_symb = stock_symb,rss_url=rss_url, chart_data_all_fth = chart_data_all_fth, stock=stockname, pred_Bayesian=returnBayesianPrice,pred_SVM=returnSVM )
else:
return render_template("stockselect.html", display='false',chart_title=stockname,
chart_data_all=chart_data_all, chart_data_all_sec=chart_data_all_sec,
chart_data_all_th=chart_data_all_th, data_types=data_types,
window=window, window2= window2, rt_price= rt_price1,
pred_price = pred_price, rec_BS = rec_BS, pred_price1 = pred_price1, rec_BS1 = rec_BS1, stock_symb = stock_symb,rss_url=rss_url, chart_data_all_fth = chart_data_all_fth, stock=stockname, pred_Bayesian=returnBayesianPrice,pred_SVM=returnSVM )
def stock():
data_types = [
'High',
'Low',
'Close',
'Volume',
'Adj Close',
]
stockname = request.args.get('stock_name')
window = 50
window2 = 150
recent_trend_query = """
SELECT Date, `{__value_name__}`
FROM seproject.historicaldata
where sym = '{__stockname__}'
order by Date ASC;
"""
move_avg_query = """
SELECT seproject.historicaldata.Date, seproject.historicaldata.`{__value_name__}`, avg(historicaldata_past.`{__value_name__}`) as `{__value_name__}_window`
FROM seproject.historicaldata
JOIN (
SELECT
seproject.historicaldata.Date, seproject.historicaldata.`{__value_name__}`
FROM seproject.historicaldata
WHERE seproject.historicaldata.sym = '{__stockname__}'
) AS historicaldata_past
ON seproject.historicaldata.Date BETWEEN historicaldata_past.Date and date_add(historicaldata_past.Date, interval + {__window__} day)
WHERE seproject.historicaldata.sym = '{__stockname__}'
GROUP BY 1, 2
order by seproject.historicaldata.Date ASC;
"""
rt_price = """
SELECT price
FROM seproject.realtimedata
WHERE sym = '{__stockname__}'
"""
#rt_price1 = {}
pred_price = round(analyzeSymbol(stockname), 2)
q_results = db.engine.execute(rt_price.format(__stockname__=stockname))
#d = json.dumps(q_results)
for res in q_results:
chart_data = res[0]
rt_price1 = chart_data
print(type(pred_price))
print(type(rt_price1))
rec_BS_A = ['BUY', 'SELL']
if float(rt_price1) < pred_price:
rec_BS = rec_BS_A[0]
else:
rec_BS = rec_BS_A[1]
chart_data_all = {}
chart_data_all_sec = {}
chart_data_all_th = {}
for data_type in data_types:
q_results = db.engine.execute(
recent_trend_query.format(__stockname__=stockname,
__value_name__=data_type))
chart_values = []
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[1]])
chart_data_all[data_type] = chart_values
q_results = db.engine.execute(
move_avg_query.format(__stockname__=stockname,
__value_name__=data_type,
__window__=window))
chart_values = []
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[2]])
chart_data_all_sec[data_type] = chart_values
q_results = db.engine.execute(
move_avg_query.format(__stockname__=stockname,
__value_name__=data_type,
__window__=window2))
chart_values = []
for result in q_results:
unixtime = time.mktime(result[0].timetuple()) * 1000
chart_values.append([unixtime, result[2]])
chart_data_all_th[data_type] = chart_values
return render_template("stockselect.html",
chart_title=stockname,
chart_data_all=chart_data_all,
chart_data_all_sec=chart_data_all_sec,
chart_data_all_th=chart_data_all_th,
data_types=data_types,
window=window,
window2=window2,
rt_price=rt_price1,
pred_price=pred_price,
rec_BS=rec_BS,
stock=stockname)