def ema(ticker,
rolling_windows=[20, 50],
hide_price=False,
graph_path=os.path.join(os.getcwd(), "ema")):
data = stock_data.get_dataframe_by_ticker(ticker)
exp_means = [
data['Adj Close'].ewm(span=i, adjust=False).mean()
for i in rolling_windows
]
for i, exp_mean in enumerate(exp_means):
plt.plot(
data['Date'],
exp_mean,
label=f'{ticker} {rolling_windows[i]} Exponential Moving Average')
if hide_price is False:
plt.plot(data['Date'],
data['Adj Close'],
label=f'{ticker} {max(rolling_windows)} Price',
color='magenta')
plt.legend(loc='upper left')
try:
plt.savefig(os.path.join(
graph_path,
f'{ticker}_{datetime.datetime.now().date().isoformat()}.png'),
format="png")
except FileNotFoundError:
os.makedirs(graph_path)
plt.savefig(os.path.join(
graph_path,
f'{ticker}_{datetime.datetime.now().date().isoformat()}.png'),
format="png")
def graph_macd(ticker,
windows=[12, 26],
signal_line=9,
graph_path=os.path.join(os.getcwd(), "macd")):
data = stock_data.get_dataframe_by_ticker(ticker)
if data is None or len(data) < min(windows):
return None
data[f'{windows[0]} ema'] = data['Adj Close'].ewm(span=windows[0],
adjust=False,
min_periods=1).mean()
data[f'{windows[1]} ema'] = data['Adj Close'].ewm(span=windows[1],
adjust=False,
min_periods=1).mean()
data['MACD'] = data[f'{windows[0]} ema'] - data[f'{windows[1]} ema']
data[f'signal line'] = data['MACD'].ewm(span=signal_line,
adjust=False).mean()
fig, ax = plt.subplots(1)
ax.plot(data['Date'][-30:], data['MACD'][-30:], label=f'MACD')
ax.plot(data['Date'][-30:],
data['signal line'][-30:],
label=f'Signal Line')
ax.axhline(0, linestyle='--')
ax.legend(loc='upper left')
try:
plt.savefig(os.path.join(
graph_path,
f'{ticker}_{datetime.datetime.now().date().isoformat()}.png'),
format='png')
except FileNotFoundError:
os.makedirs(graph_path)
plt.savefig(os.path.join(
graph_path,
f'{ticker}_{datetime.datetime.now().date().isoformat()}.png'),
format='png')
def get_rsi_dataframe(ticker,
start=two_years_from_today,
end=today,
time_window=14):
df = stock_data.get_dataframe_by_ticker(ticker, start, end)
if df is not None:
df['RSI'] = compute_rsi(df['Adj Close'], time_window)
return df
def get_macd_intersection(ticker,
windows=[12, 26],
signal_line=9,
look_back_time=5):
data = stock_data.get_dataframe_by_ticker(ticker)
if data is None or len(data) < min(windows):
return None
data[f'{windows[0]} ema'] = data['Adj Close'].ewm(span=windows[0],
adjust=False,
min_periods=1).mean()
data[f'{windows[1]} ema'] = data['Adj Close'].ewm(span=windows[1],
adjust=False,
min_periods=1).mean()
data['MACD'] = data[f'{windows[0]} ema'] - data[f'{windows[1]} ema']
data[f'signal line'] = data['MACD'].ewm(span=signal_line,
adjust=False).mean()
intersections = {}
for i in range(len(data['Date'][-look_back_time:])):
if len(data['Date'][:]) - look_back_time + 1 + i < len(
data['Date'][:]):
y11 = data[f'MACD'][len(data['Date'][:]) - look_back_time +
i] # macd y1
y21 = data[f'signal line'][len(data['Date'][:]) - look_back_time +
i] # signal y1
y12 = data[f'MACD'][len(data['Date'][:]) - look_back_time + 1 +
i] #macd y2
y22 = data[f'signal line'][len(data['Date'][:]) - look_back_time +
1 + i] # signal y2
if y11 > y21 and y12 < y22:
intersections = {
"ticker":
ticker,
"cross_date":
data['Date'][len(data['Date'][:]) - look_back_time + 1 +
i],
"sentiment":
"Bearish",
"count":
-1
}
if y21 > y11 and y22 < y12:
intersections = {
"ticker":
ticker,
"cross_date":
data['Date'][len(data['Date'][:]) - look_back_time + 1 +
i],
"sentiment":
"Bullish",
"count":
1
}
return intersections
def get_sma_intersection(ticker, windows=[20, 50], look_back_time=5):
data = stock_data.get_dataframe_by_ticker(ticker)
if data is None or len(data) < min(windows):
return None
data[f'{windows[0]} SMA'] = data['Adj Close'].rolling(
window=windows[0]).mean()
data[f'{windows[1]} SMA'] = data['Adj Close'].rolling(
window=windows[1]).mean()
intersections = {}
for i in range(len(data['Date'][-look_back_time:])):
if len(data['Date'][:]) - look_back_time + 1 + i < len(
data['Date'][:]):
y11 = data[f'{windows[0]} SMA'][len(data['Date'][:]) -
look_back_time +
i] # short window n day price
y21 = data[f'{windows[1]} SMA'][len(data['Date'][:]) -
look_back_time +
i] # long window n day price
y12 = data[f'{windows[0]} SMA'][len(data['Date'][:]) -
look_back_time + 1 +
i] # short window n + 1 price
y22 = data[f'{windows[1]} SMA'][len(data['Date'][:]) -
look_back_time + 1 +
i] # long window n + 1 price
if y11 > y21 and y12 < y22:
intersections = {
"ticker":
ticker,
"cross_date":
data['Date'][len(data['Date'][:]) - look_back_time + 1 +
i],
"sentiment":
"Bearish",
"count":
-1
}
if y21 > y11 and y22 < y12:
intersections = {
"ticker":
ticker,
"cross_date":
data['Date'][len(data['Date'][:]) - look_back_time + 1 +
i],
"sentiment":
"Bullish",
"count":
1
}
return intersections
def sma(ticker,
rolling_windows=[20, 50],
hide_price=False,
bollinger=False,
graph_path=os.path.join(os.getcwd(), "sma")):
data = stock_data.get_dataframe_by_ticker(ticker)[-2 *
(max(rolling_windows)):]
rolling_means = [
data['Adj Close'].rolling(window=i).mean() for i in rolling_windows
]
for i, rolling_mean in enumerate(rolling_means):
plt.plot(data['Date'],
rolling_mean,
label=f'{ticker} {rolling_windows[i]} Simple Moving Average')
if bollinger is True:
data = bollinger_bands(data, window=rolling_windows[0])
plt.plot(data['Date'],
data['Lower Bollinger'],
label=f'{ticker} {rolling_windows[0]} Lower Bollinger Band')
plt.plot(data['Date'],
data['Upper Bollinger'],
label=f'{ticker} {rolling_windows[0]} Upper Bollinger Band')
if hide_price is False:
plt.plot(data['Date'],
data['Adj Close'],
label=f'{ticker} {max(rolling_windows)} Price',
color='magenta')
plt.legend(loc='upper left')
try:
plt.savefig(os.path.join(
graph_path,
f'{ticker}_{datetime.datetime.now().date().isoformat()}.png'),
format='png')
except FileNotFoundError:
os.makedirs(graph_path)
plt.savefig(os.path.join(
graph_path,
f'{ticker}_{datetime.datetime.now().date().isoformat()}.png'),
format='png')
def get_rsi_plot(ticker,
start=two_years_from_today,
end=today,
time_window=14):
df = stock_data.get_dataframe_by_ticker(ticker, start, end)
df['RSI'] = compute_rsi(df['Adj Close'], time_window)
fig, (ax1, ax2) = plt.subplots(2)
fig.set_figwidth(15)
fig.set_figheight(5)
fig.tight_layout(pad=3.0)
ax1.set_title("Time vs Price")
ax1.set
ax1.plot(df['Date'], df['Adj Close'])
ax2.set_title('RSI chart')
ax2.plot(df['Date'], df['RSI'])
ax2.axhline(0, linestyle='--', alpha=0.1)
ax2.axhline(20, linestyle='--', alpha=0.5)
ax2.axhline(30, linestyle='--')
ax2.axhline(70, linestyle='--')
ax2.axhline(80, linestyle='--', alpha=0.5)
ax2.axhline(100, linestyle='--', alpha=0.1)
#plt.savefig(f'{ticker}_RSI.png')
plt.show()
def get_volume(ticker, start=fourteen_days_ago, end=stock_data.today):
data = stock_data.get_dataframe_by_ticker(ticker, start_time=fourteen_days_ago, end_time=stock_data.today)
plt.plot(data['Date'], data['Volume'])
plt.show()