def analyze(self):
'''
Analyze each of the Trade objects in TradeList depending
on value of settings.get('counter', 'strats') and add the
calculated attributes to Trade
:returns
TradeList
'''
#these are the strategies that will be analysed using the Counter pattern
strats = self.settings.get('counter', 'strats').split(",")
trade_list = []
for t in self.tlist:
tl_logger.info("Processing trade: {0}-{1}".format(t.pair, t.start))
if t.strat in strats:
if t.entered is False and not getattr(t, 'outcome'):
t.run_trade()
c = Counter(trade=t,
settingf=self.settingf,
settings=self.settings,
init_feats=True)
attrb_ls = self.settings.get('counter', 'attrbs').split(",")
for a in attrb_ls:
# add 'a' attribute to Trade object
setattr(t, a, getattr(c, a))
trade_list.append(t)
tl_logger.info("Done")
tl = TradeList(settingf=self.settingf,
settings=self.settings,
tlist=trade_list)
return tl
def test_set_pivots_lasttime(pair, id, start, type, SR, SL, TP, entry, dates,
clean_tmp):
"""
Check that self.pivots_lasttime class attribute
has been initialized
"""
t = Trade(id=id,
start=start,
pair=pair,
timeframe='D',
type=type,
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, period=1000, settingf='../../data/settings.ini')
c.set_lasttime()
c.set_pivots()
c.set_pivots_lasttime()
times = []
for p in c.pivots_lasttime.plist:
times.append(p.candle.time)
assert dates == times
def test_set_pivots_4h(pair, timeframe, id, start, type, SR, SL, TP, entry,
pllen, settings_obj, clean_tmp):
"""
Check that self.pivots class attribute is correctly set using a H4 timeframe
"""
settings_obj.set('pivots', 'th_bounces', '0.01')
t = Trade(id=id,
start=start,
pair=pair,
timeframe=timeframe,
type=type,
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settings=settings_obj)
c = Counter(trade=t, settings=settings_obj)
c.set_pivots()
assert len(c.pivots.plist) == pllen
def analyze(self):
'''
Analyze each of the Trade objects in TradeList depending
on value of settings.get('counter', 'strats') and add the
calculated attributes to Trade
:returns
TradeList
'''
#these are the strategies that will be analysed using the Counter pattern
tl_logger.info("Strategies that will be analysed: {0}".format(
self.settings.get('counter', 'strats')))
strats = self.settings.get('counter', 'strats').split(",")
trade_list = []
for t in self.tlist:
tl_logger.info("Processing trade: {0}-{1}".format(t.pair, t.start))
if t.strat in strats:
if t.entered is False and (not hasattr(t, 'outcome')
or math.isnan(t.outcome) is True):
t.run_trade()
c = Counter(trade=t,
settingf=self.settingf,
settings=self.settings,
ser_data_obj=self.ser_data_obj,
init_feats=True)
tl_logger.debug("Counter attributes analysed:{0}".format(
self.settings.get('counter', 'attrbs').split(",")))
attrb_ls = self.settings.get('counter', 'attrbs').split(",")
for a in attrb_ls:
if hasattr(c, a) is True:
# add 'a' attribute to Trade object
setattr(t, a, getattr(c, a))
else:
tl_logger.warn(
"Attribute {0} is not defined in Counter object. Skipping..."
.format(a))
setattr(t, a, "n.a.")
else:
tl_logger.debug(
"Trade.strat ({0}) is not within list of trades to analyse. Skipping..."
.format(t.strat))
trade_list.append(t)
tl_logger.info("Done")
tl = TradeList(settingf=self.settingf,
settings=self.settings,
tlist=trade_list)
return tl
def ct_object():
'''Returns Counter object'''
t = Trade(id='EUR_GBP_13AUG2019D',
start='2019-08-12 22:00:00',
pair='EUR_GBP',
timeframe='D',
type='short',
SR=0.92909,
SL=0.93298,
TP=0.90366,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, settingf='../../data/settings.ini')
return c
def test_max_min_rsi(pair, timeframe, id, start, type, SR, SL, TP, entry,
avalue):
t = Trade(id=id,
start=start,
pair=pair,
timeframe=timeframe,
type=type,
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, settingf='../../data/settings.ini')
assert avalue == c.max_min_rsi()
def test_is_entry_onrsi(pair, id, timeframe, start, type, SR, SL, TP, entry,
entry_onrsi, clean_tmp):
t = Trade(id=id,
start=start,
pair=pair,
timeframe=timeframe,
type=type,
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, settingf='../../data/settings.ini', init_feats=True)
assert entry_onrsi == c.is_entry_onrsi()
def test_set_pips_c_trend(pair, id, timeframe, start, type, SR, SL, TP, entry,
pips_c_trend, clean_tmp):
t = Trade(id=id,
start=start,
pair=pair,
timeframe=timeframe,
type=type,
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, settingf='../../data/settings.ini', init_feats=True)
assert pips_c_trend == c.pips_c_trend
def test_set_trend_i(pair, id, timeframe, start, type, SR, SL, TP, entry,
trend_i, clean_tmp):
t = Trade(id=id,
start=start,
pair=pair,
timeframe=timeframe,
type=type,
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, settingf='../../data/settings.ini')
c.set_trend_i()
assert trend_i == c.trend_i
def test_ctobject_notype(start, SR, SL, TP, entry, clean_tmp):
"""
Check that Counter object without
a type defined is correctly instantiated
"""
t = Trade(id='test',
start=start,
pair='EUR_AUD',
timeframe='D',
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, period=1000, settingf='../../data/settings.ini')
assert 0
def test_set_lasttime(start, type, SR, SL, TP, entry, lasttime, clean_tmp):
"""
Check that self.lasttime class attribute
has been initialized
"""
t = Trade(id='test',
start=start,
pair='EUR_AUD',
timeframe='D',
type=type,
SR=SR,
SL=SL,
TP=TP,
entry=entry,
strat='counter_b1',
settingf="../../data/settings.ini")
c = Counter(trade=t, period=1000, settingf='../../data/settings.ini')
c.set_lasttime()
assert c.lasttime == lasttime
def calc_SR(self, adateObj):
'''
Function to calculate S/R lines
Parameters
----------
datetime object used for identifying
S/R areas
Return
------
HAreaList object
'''
# calculate price range for calculating S/R
ul, ll = self.get_max_min(adateObj)
tb_logger.info("Running calc_SR for estimated range: {0}-{1}".format(
ll, ul))
prices = []
bounces = [] # contains the number of pivots per price level
score_per_bounce = []
tot_score = []
# the increment of price in number of pips is double the hr_extension
prev_p = None
p = float(ll)
while p <= float(ul):
tb_logger.debug("Processing S/R at {0}".format(round(p, 4)))
# each of 'p' will become a S/R that will be tested for bounces
# set entry to price+self.settings.getint('trade_bot','i_pips')
entry = add_pips2price(self.pair, p,
self.settings.getint('trade_bot', 'i_pips'))
# set S/L to price-self.settings.getint('trade_bot','i_pips')
SL = substract_pips2price(
self.pair, p, self.settings.getint('trade_bot', 'i_pips'))
t = Trade(id='{0}.{1}.detect_sr.{2}'.format(
self.pair, adateObj.isoformat(), round(p, 5)),
start=adateObj.strftime('%Y-%m-%d %H:%M:%S'),
pair=self.pair,
timeframe=self.timeframe,
type='short',
entry=entry,
SR=p,
SL=SL,
RR=1.5,
strat='counter_b1',
settingf=self.settingf,
settings=self.settings)
# reduce period that Counter uses so the S/R
# are calculated with the most recent pivots
newperiod = self.settings.get('trade_bot', 'period_range')
self.settings.set('counter', 'period', newperiod)
c = Counter(trade=t, init_feats=True, settings=self.settings)
if len(c.pivots.plist) == 0:
mean_pivot = 0
else:
mean_pivot = round(c.score_pivot, 2)
prices.append(round(p, 5))
bounces.append(len(c.pivots.plist))
tot_score.append(c.total_score)
score_per_bounce.append(mean_pivot)
# increment price to following price.
# Because the increment is made in pips
# it does not suffer of the JPY pairs
# issue
p = add_pips2price(self.pair, p,
2 * self.settings.getint('trade_bot', 'i_pips'))
if prev_p is None:
prev_p = p
else:
increment_price = round(p - prev_p, 5)
prev_p = p
data = {
'price': prices,
'bounces': bounces,
'scores': score_per_bounce,
'tot_score': tot_score
}
df = pd.DataFrame(data=data)
### establishing bounces threshold as the args.th quantile
# selecting only rows with at least one pivot and tot_score>0,
# so threshold selection considers only these rows
# and selection is not biased when range of prices is wide
dfgt1 = df.loc[(df['bounces'] > 0)]
dfgt2 = df.loc[(df['tot_score'] > 0)]
bounce_th = dfgt1.bounces.quantile(
self.settings.getfloat('trade_bot', 'th'))
score_th = dfgt2.tot_score.quantile(
self.settings.getfloat('trade_bot', 'th'))
print("Selected number of pivot threshold: {0}".format(bounce_th))
print("Selected tot score threshold: {0}".format(score_th))
# selecting records over threshold
dfsel = df.loc[(df['bounces'] > bounce_th) |
(df['tot_score'] > score_th)]
# repeat until no overlap between prices
ret = self.__calc_diff(dfsel, increment_price)
dfsel = ret[0]
tog_seen = ret[1]
while tog_seen is True:
ret = self.__calc_diff(dfsel, increment_price)
dfsel = ret[0]
tog_seen = ret[1]
# iterate over DF with selected SR to create a HAreaList
halist = []
for index, row in dfsel.iterrows():
resist = HArea(price=row['price'],
pips=self.settings.getint('pivots', 'hr_pips'),
instrument=self.pair,
granularity=self.timeframe,
no_pivots=row['bounces'],
tot_score=row['tot_score'],
settingf=self.settingf,
settings=self.settings)
halist.append(resist)
halist = HAreaList(halist=halist,
settingf="data/settings.ini",
settings=self.settings)
tb_logger.info("Run done")
return halist
entry = add_pips2price(args.instrument, p, 30)
# set S/L to price-30pips
SL = substract_pips2price(args.instrument, p, 30)
t = Trade(id='{0}.detect_sr.{1}'.format(args.instrument, round(p, 5)),
start=args.start,
pair=args.instrument,
timeframe='D',
type='short',
entry=entry,
SR=p,
SL=SL,
RR=1.5,
strat='counter_b1',
settingf=args.settingf)
c = Counter(trade=t, init_feats=True, settingf=args.settingf)
ratio = None
if len(c.pivots.plist) == 0:
ratio = 0
mean_pivot = 0
else:
mean_pivot = round(c.score_pivot, 2)
file.write("{0}\t{1}\t{2}\t{3}\n".format(round(p, 5), len(c.pivots.plist),
c.total_score, mean_pivot))
prices.append(round(p, 5))
bounces.append(len(c.pivots.plist))
tot_scores.append(c.total_score)
score_per_bounce.append(mean_pivot)
p = add_pips2price(args.instrument, p, 60)
