def get_stock(stock_name):
""" render template for stock page of site. USe Jinja to dynamically display assignment data
:return: HTML rendering
"""
# look up stock in mongo db
stock = mongo.db.stocks.find_one({"_id": stock_name})
# check that a result was returned from db
if stock is None:
abort(404)
# tack on plotly urls
stock['price_plot'] = plot_stock_price(stock)
stock['boxplot'] = plot_returns_boxplot_stock(stock)
stock['returns_plot'] = plot_yearly_returns(stock)
# tack on statistics
stock['max'] = calculate_historical_max(stock)
stock['min'] = calculate_historical_min(stock)
stock['best'] = calculate_largest_daily_gain(stock)
stock['worst'] = calculate_largest_daily_loss(stock)
# predict future price using ML
stock_data = list(mongo.db.stocks.find())
features, targets = extract_feature_data(stock_data)
# create the svm and dtree
svm = create_support_vector_regression(features, targets)
dtree = create_decision_tree_regression(features, targets)
# predict price of the stock
features_stock, _ = extract_feature_data([stock])
stock['future_returns'] = (svm.predict(features_stock)[0] +
dtree.predict(features_stock)[0]) / 2
# render the HTML
return render_template('stock.html', stock=stock)
def test_svm_regression(self):
# extract features and labels
features, labels = extract_feature_data([self.test_stock, self.test_stock2])
# create decision tree classifier
svm = create_support_vector_regression(features, labels)
# check manually set parameters of svm correct
self.assertEqual(svm.kernel, 'linear')
self.assertEqual(svm.C, 100)
def test_dtree_regression(self):
# extract features and labels
features, labels = extract_feature_data([self.test_stock, self.test_stock2])
# create decision tree classifier
dtree = create_decision_tree_regression(features, labels)
# check manually set parameters of decision tree correct
self.assertEqual(dtree.max_depth, 2)
self.assertEqual(dtree.min_samples_leaf, 1)
def test_extra_feature_data_single(self):
# extract the features and labels
features, labels = extract_feature_data([self.test_stock2])
# check that features and labels correctly created
self.assertEqual(len(features), 1)
self.assertIn([0.0], features)
self.assertEqual(len(labels), 1)
self.assertIn(-40.938865746148089, list(labels))
def get_portfolio(portfolio_name):
""" render template for portfolio page of site. USe Jinja to dynamically display assignment data
:return: HTML rendering
"""
# look up portfolio in mongo db
portfolio = mongo.db.portfolios.find_one({"_id": portfolio_name})
index_data = mongo.db.stocks.find_one({"_id": "^GSPC"})
# check that a result was returned from db
if portfolio is None or index_data is None:
abort(404)
# tack on value field
portfolio['value'] = get_total_value_portfolio(portfolio)
# tack on the plotly urls
portfolio['breakdown_plot'] = plot_portfolio_breakdown(portfolio)
portfolio['value_plot'] = plot_portfolio_value_vs_time(portfolio)
portfolio['returns_plot'] = plot_portfolio_returns_vs_market(
portfolio, index_data)
portfolio['boxplot'] = plot_returns_boxplot(portfolio, index_data)
# tack on statistics
portfolio['alpha'] = get_portfolio_alpha(portfolio, index_data)
portfolio['beta'] = get_portfolio_beta(portfolio, index_data)
portfolio['sharpe_ratio'] = get_portfolio_sharpe_ratio(
portfolio, index_data)
portfolio['jensen_measure'] = get_portfolio_jensen_measure(
portfolio, index_data)
# add stock data
stock_data = list(mongo.db.stocks.find())
# perform ML
features, targets = extract_feature_data(stock_data)
svm = create_support_vector_regression(features, targets)
dtree = create_decision_tree_regression(features, targets)
# add future return fields
stocks = {}
for stock in stock_data:
features_stock, _ = extract_feature_data([stock])
if len(features_stock) == 0:
stocks[stock['symbol']] = 0.0
else:
stocks[
stock['symbol']] = (svm.predict(features_stock)[0] +
dtree.predict(features_stock)[0]) / 2
return render_template('portfolio.html',
portfolio=portfolio,
stocks=stocks)
def test_extra_feature_data_multiple(self):
# extract the features and labels
features, labels = extract_feature_data([self.test_stock, self.test_stock2])
# check that features and labels correctly created
self.assertEqual(len(features), 2)
self.assertIn([0.0], features)
self.assertIn([1.2], features)
self.assertEqual(len(labels), 2)
self.assertIn(-40.938865746148089, list(labels))
self.assertIn(-3.6179451381311551, list(labels))
def get_stocks():
""" render template for stock page of site. USe Jinja to dynamically display assignment data
:return: HTML rendering
"""
# look up stock in mongo db
stock_data = list(mongo.db.stocks.find())
# check that a result was returned from db
if stock_data is None:
abort(404)
# perform ML algos
features, targets = extract_feature_data(stock_data)
# create the svm and dtree
svm = create_support_vector_regression(features, targets)
dtree = create_decision_tree_regression(features, targets)
# tack on plotly urls
formatted_data = format_stock_data(stock_data)
clusters = k_means(formatted_data, 3)
cluster_assignments = get_cluster_assignments(formatted_data, clusters)
plots = {}
plots['kmeans'] = plot_clusters(formatted_data, clusters)
plots['hierarchy'] = plot_hierarchy(formatted_data)
plots['svm'] = plot_support_vector_regression(features, targets, svm)
plots['dtree'] = plot_decision_tree_regression(features, targets,
dtree)
# add expected returns
stock_rets = {}
for stock in stock_data:
features_stock, _ = extract_feature_data([stock])
if len(features_stock) == 0:
stock_rets[stock['symbol']] = 0.0
else:
stock_rets[
stock['symbol']] = (svm.predict(features_stock)[0] +
dtree.predict(features_stock)[0]) / 2
# render the HTML
return render_template('stocks.html',
stocks=stock_data,
plots=plots,
clusters=cluster_assignments,
stock_rets=stock_rets)
def get_svm():
""" API endpoint to get regression line for stock performance
:param symbol: stock to do regression for
:return: JSON representation regression line
"""
# get stock data from db
stock_data = mongo.db.stocks.find()
if stock_data is None:
abort(404)
# format the data
features, targets = extract_feature_data(list(stock_data))
# create the svm
svm = create_support_vector_regression(features, targets)
# make the plot
get_accuracy(features, targets, svm.predict(features))
res = plot_support_vector_regression(features, targets, svm)
return jsonify({"plot_url": res}), 200