def slinger(ax, datafile, ticker, parameters):
time = datafile[ticker]['datetime'][...]
data_open = datafile[ticker]['open'][...]
data_high = datafile[ticker]['high'][...]
data_low = datafile[ticker]['low'][...]
data_volume = datafile[ticker]['volume'][...]
datafile.close()
tradeable = Analytics.market_hours(t=time)
print(np.sum(tradeable))
candle = Analytics.candle_avg(open=data_open, high=data_high, low=data_low)
candle_low_bollinger, candle_high_bollinger = Analytics.candle_bollinger_bands(
open=data_open,
high=data_high,
low=data_low,
average=candle,
period=30)
period = 30
sma = Analytics.moving_average(data=candle, period=period)
sma_short = Analytics.moving_average(data=candle, period=period // 3)
sma_low_bollinger, sma_high_bollinger = Analytics.bollinger_bands(
data=sma_short, average=sma)
sma_d = Analytics.derivative(sma, period=period // 6)
# sma_d = Analytics.moving_average(sma_d, period=period // 6)
sma_dd = Analytics.second_derivative(sma, period=period)
day_volatility = Analytics.day_volatility(data=candle, tradeable=tradeable)
print('day volatility: {}'.format(day_volatility))
# find the best strategy of given strategies:
performance_list = []
# parameters['option_type'] = position_class.OptionType.PUT
for parameter in parameters:
parameter['VIX'] = day_volatility
results_list = PutSlingerBollinger.Bollinger_strat(
time=time,
sma=sma,
sma_short=sma_short,
bollinger_up=sma_high_bollinger,
bollinger_down=sma_low_bollinger,
sma_d=sma_d,
candle=candle,
candle_high=candle_high_bollinger,
candle_low=candle_low_bollinger,
parameters=parameter)
put_buy_option_price = results_list[2]
put_sell_option_price = results_list[5]
put_percent = (put_sell_option_price -
put_buy_option_price) / put_buy_option_price
put_percent[
put_percent > 5] = 5 # put an upper bound on the option returns.
put_percent_avg = np.sum(put_percent) / put_percent.shape[0]
performance_list.append(put_percent_avg)
performance_array = np.array(performance_list)
print('performance_array: {}'.format(performance_array))
best_perf_loc = np.where(
performance_array == performance_array.max())[0][0]
# hack the parameters to be the best choice for the rest of the function, this keeps recoding to a minimum:
parameter = parameters[best_perf_loc]
parameter = parameters[0]
results_list = PutSlingerBollinger.Bollinger_strat(
time=time,
sma=sma,
sma_short=sma_short,
bollinger_up=sma_high_bollinger,
bollinger_down=sma_low_bollinger,
sma_d=sma_d,
candle=candle,
candle_high=candle_high_bollinger,
candle_low=candle_low_bollinger,
parameters=parameter)
put_buy_locs = results_list[0]
put_buy_price = results_list[1]
put_buy_option_price = results_list[2]
put_sell_locs = results_list[3]
put_sell_price = results_list[4]
put_sell_option_price = results_list[5]
position_value = results_list[7]
print('stock price at open: {}'.format(put_buy_price))
print('strike price: {}'.format(results_list[6]))
print('stock price at close: {}'.format(put_sell_price))
print('option cost at open: {}'.format(put_buy_option_price))
print('option cost at close: {}'.format(put_sell_option_price))
put_profits = (put_sell_option_price - put_buy_option_price)
put_percent = (put_sell_option_price -
put_buy_option_price) / put_buy_option_price
print('option % gain: {}'.format(put_percent))
put_percent[
put_percent > 5] = 5 # put an upper bound on the option returns.
print('position values: {}'.format(position_value))
account_value = np.sum(position_value)
print('account value ate EOD: {}'.format(account_value))
# put_percent_avg = np.sum(put_percent) / put_percent.shape[0]
put_percent_avg = np.sum(account_value) - 1
print('average option % gain: {}'.format(put_percent_avg))
results_list.append(put_percent)
results_list.append(put_percent_avg)
focus_top = time.shape[0] - 60 * 48
focus_bot = time.shape[0] + 1
focus_top = 0
focus_bot = time.shape[0] + 1
candle_rescaled = candle - np.sum(candle) / sma.shape[0]
candle_rescaled = candle_rescaled / np.abs(candle_rescaled).max()
sma_rescaled = sma - np.sum(sma) / sma.shape[0]
sma_rescaled = sma_rescaled / np.abs(sma_rescaled).max()
Bollinger_oscillator = 2 * (
sma_short - sma) / np.absolute(sma_high_bollinger - sma_low_bollinger)
minute_time = Analytics.minute_time(time)
# print(minute_time)
'''
ax[0].plot(time[focus_top:focus_bot], candle_rescaled, label='candle')
ax[0].plot(time[focus_top:focus_bot], sma_rescaled, label='sma')
ax[0].plot(time[focus_top:focus_bot], sma_d[focus_top:focus_bot] / np.abs(sma_d).max(), label='sma_d')
ax[0].plot(time[focus_top:focus_bot], sma_dd[focus_top:focus_bot] / np.abs(sma_dd).max(), label='sma_dd')
ax[0].legend()
'''
#################################################################################
# plt.figure(figsize=(20, 10))
# plt.suptitle('profitable trades')
# ax[0].plot(time[tradeable], data_volume[tradeable], '.')
ax_twin = ax[0].twinx()
ax_twin.plot(minute_time[tradeable], Bollinger_oscillator[tradeable])
ax_twin.plot(minute_time[tradeable],
np.ones_like(minute_time[tradeable]) *
parameter['Bollinger_top'],
color='k')
ax_twin.plot(minute_time[tradeable],
np.ones_like(minute_time[tradeable]) *
parameter['Bollinger_bot'],
color='k')
ax[0].plot(minute_time[tradeable],
candle[tradeable],
'.',
label=str(put_percent_avg))
ax[0].plot(minute_time[tradeable], sma[tradeable])
ax[0].plot(minute_time[tradeable], sma_low_bollinger[tradeable])
ax[0].plot(minute_time[tradeable], sma_high_bollinger[tradeable])
ax[0].plot(minute_time[tradeable], candle_low_bollinger[tradeable])
ax[0].plot(minute_time[tradeable], candle_high_bollinger[tradeable])
profit_put_buy_locs = put_buy_locs[put_profits >= 0]
put_cut = profit_put_buy_locs[profit_put_buy_locs > focus_top]
ax[0].plot(minute_time[put_cut], candle[put_cut], '>', color='k')
profit_put_sell_locs = put_sell_locs[put_profits >= 0]
put_cut = profit_put_sell_locs[profit_put_sell_locs > focus_top]
ax[0].plot(minute_time[put_cut], candle[put_cut], '<', color='k')
ax[0].legend()
#################################################################################
# plt.figure(figsize=(20, 10))
# plt.suptitle('loss trades')
# ax[1].plot(minute_time, data_volume, '.')
ax_twin = ax[1].twinx()
ax_twin.plot(minute_time[tradeable], Bollinger_oscillator[tradeable])
ax_twin.plot(minute_time[tradeable],
np.ones_like(minute_time[tradeable]) *
parameter['Bollinger_top'],
color='k')
ax_twin.plot(minute_time[tradeable],
np.ones_like(minute_time[tradeable]) *
parameter['Bollinger_bot'],
color='k')
ax[1].plot(minute_time[tradeable],
candle[tradeable],
'.',
label=str(put_percent_avg))
ax[1].plot(minute_time[tradeable], sma[tradeable])
ax[1].plot(minute_time[tradeable], sma_low_bollinger[tradeable])
ax[1].plot(minute_time[tradeable], sma_high_bollinger[tradeable])
ax[1].plot(minute_time[tradeable], candle_low_bollinger[tradeable])
ax[1].plot(minute_time[tradeable], candle_high_bollinger[tradeable])
loss_put_buy_locs = put_buy_locs[put_profits < 0]
put_cut = loss_put_buy_locs[loss_put_buy_locs > focus_top]
ax[1].plot(minute_time[put_cut], candle[put_cut], '>', color='k')
loss_put_sell_locs = put_sell_locs[put_profits < 0]
put_cut = loss_put_sell_locs[loss_put_sell_locs > focus_top]
ax[1].plot(minute_time[put_cut], candle[put_cut], '<', color='k')
ax[1].legend()
return put_percent_avg, performance_array
print('total put profits: {}'.format(np.sum(put_profits)))
put_percent = (put_sell_option_price - put_buy_option_price) / put_buy_option_price
print('put percents: {}'.format(put_percent))
print('total put percent: {}'.format(np.sum(put_percent) / put_percent.shape[0]))
focus_top = 0
focus_bot = time.shape[0] + 1
candle_rescaled = candle - np.sum(candle) / sma.shape[0]
candle_rescaled = candle_rescaled / np.abs(candle_rescaled).max()
sma_rescaled = sma - np.sum(sma) / sma.shape[0]
sma_rescaled = sma_rescaled / np.abs(sma_rescaled).max()
Bollinger_oscillator = 2 * (sma_short - sma) / np.absolute(sma_high_bollinger - sma_low_bollinger)
minute_time = Analytics.minute_time(time)
print('put buy locs: {}'.format(minute_time[put_buy_locs]))
print('put sell locs: {}'.format(minute_time[put_sell_locs]))
#print(minute_time)
#################################################################################
# plt.figure(figsize=(20, 10))
# plt.suptitle('profitable trades')
# ax[0].plot(time[tradeable], data_volume[tradeable], '.')
fig, ax = plt.subplots(nrows=1, ncols=2, sharex=False, figsize=(20, 8))
ax_twin = ax[0].twinx()
ax_twin.plot(minute_time[tradeable], Bollinger_oscillator[tradeable])
ax_twin.plot(minute_time[tradeable], np.ones_like(minute_time[tradeable]) * parameter['Bollinger_top'], color='k')
ax_twin.plot(minute_time[tradeable], np.ones_like(minute_time[tradeable]) * parameter['Bollinger_bot'], color='k')
