def multiple_high_volume_small_candle_event_time(ohlc_df,
lookback_volume=60,
threshold_volume=10,
threshold_candle=1.5,
count_timeframe=60,
event_count=2,
prolong_timeframe=240,
multiplier=1,
start_time="10:00",
end_time="10:30",
delay=None):
'''alpha based on multiple big tails in small count_timeframe
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
del high, low
volume_events = ev.get_high_volume_events(volume, lookback_volume,
threshold_volume)
candle_events = ev.get_small_candle_body_events(close, opens,
threshold_candle,
lookback_volume)
events_hvsc = volume_events.multiply(candle_events)
events = ev.get_event_count_events(events_hvsc, event_count,
count_timeframe)
events = ti.null_daytime_timeperiod(events, start_time, end_time)
weights = events.rolling(prolong_timeframe).apply(prolong)
if delay:
weights = ti.delay(weights, delay)
return weights * multiplier
def acceleration_event(ohlc_df,
lookback=60,
prolong_timeframe=240,
threshold=0.1):
'''alpha based of long term acceleration
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
del volume, opens, high, low
events = ev.get_acc_events(close, lookback, threshold)
weights = events.rolling(prolong_timeframe).apply(prolong)
weights.fillna(0, inplace=True)
return weights
def vwap_event(ohlc_df,
threshold=.01,
vwap_timeframe=60,
prolong_timeframe=240,
multiplier=1):
'''alpha based on close being too far from vwap
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
del opens, high, low
events = ev.get_vwap_events(close, volume, threshold, vwap_timeframe)
weights = events.rolling(prolong_timeframe).apply(prolong)
return weights * multiplier
def high_volume_small_candle_event(ohlc_df,
lookback=60,
threshold_volume=10,
threshold_candle=1.5,
prolong_timeframe=240,
multiplier=1):
'''alpha based on high volume with in the small candle
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
del high, low
volume_events = ev.get_high_volume_events(volume, lookback,
threshold_volume)
candle_events = ev.get_small_candle_body_events(close, opens,
threshold_candle, lookback)
events = volume_events.multiply(candle_events)
weights = events.rolling(prolong_timeframe).apply(prolong)
return weights * multiplier
def multiple_tails_event(ohlc_df,
threshold_tail=1.005,
count_timeframe=60,
tail_count=2,
prolong_timeframe=240,
multiplier=1,
delay=None):
'''alpha based on multiple big tails in small count_timeframe
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
del volume
candle_events = ev.get_big_tail_events(close, high, low, opens,
threshold_tail)
events = ev.get_event_count_events(candle_events, tail_count,
count_timeframe)
weights = events.rolling(prolong_timeframe).apply(prolong)
if delay:
weights = ti.delay(weights, delay)
return weights * multiplier
def high_volume_big_tail_event(ohlc_df,
lookback=60,
threshold_volume=10,
threshold_tail=1.005,
prolong_timeframe=240,
multiplier=1,
delay=None):
'''alpha based on high volume with in the big candle
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
volume_events = ev.get_high_volume_events(volume, lookback,
threshold_volume)
candle_events = ev.get_big_tail_events(close, high, low, opens,
threshold_tail)
events = volume_events.multiply(candle_events)
weights = events.rolling(prolong_timeframe).apply(prolong)
if delay:
weights = ti.delay(weights, delay)
return weights * multiplier
def multiple_vwap_event(ohlc_df,
threshold=.01,
vwap_timeframe=60,
count_timeframe=60,
event_count=2,
prolong_timeframe=240,
multiplier=1,
delay=None):
'''alpha based on multiple vwap events
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
del opens, high, low
vwap_events = ev.get_vwap_events(close, volume, threshold, vwap_timeframe)
events = ev.get_event_count_events(vwap_events, event_count,
count_timeframe)
weights = events.rolling(prolong_timeframe).apply(prolong)
if delay:
weights = ti.delay(weights, delay)
return weights * multiplier
def high_volume_vwap_event(ohlc_df,
lookback_volume=1000,
threshold_volume=10,
threshold_vwap=1.005,
vwap_timeframe=60,
prolong_timeframe=240,
multiplier=1,
delay=None):
'''alpha based on high volume and price being far from vwap
'''
volume, opens, high, low, close = unpack_ohlc(ohlc_df)
del high, low, opens
volume_events = ev.get_high_volume_events(volume, lookback_volume,
threshold_volume)
vwap_events = ev.get_vwap_events(close, volume, threshold_vwap,
vwap_timeframe)
events = volume_events.multiply(vwap_events)
weights = events.rolling(prolong_timeframe).apply(prolong)
if delay:
weights = ti.delay(weights, delay)
return weights * multiplier