Indicator with look back
J T last edited by
I am at wits end after searching for answers and also reading the dox. I reckon my understanding is still not good enough and hoping for some expert advice. My indicator is not complete yet, but stuck at some parts. I am trying to do it as declarative in __init__as it possibly can as I feel a part of me missing if I dont.
Code as follows and I keep getting an empty array in ys which I think its the problem.
class CandlePower(bt.Indicator): """ 1. Buyers or Sellers in control - e.g green candle - close price is closer to top or bottom of candle 2. Is there strength in the move - size of candle relative to the others in the period. Use a plot of xs and ys, find slope of the line through the points. xs is just the x axis 1, 2 & 3. ys are size of candle bar. """ lines = ('candle_power', 'perc_of_range', 'slope') params = (dict(period=3, safediv=True)) plotlines = dict( candle_power=dict(_name='CP', marker='v', markersize=6.0, ls='', color='pink', _skipnan=True), slope=dict(_name='slope', _plotskip=True), perc_of_range=dict(_name='perc_of_range', _plotskip=True), ) plotinfo = dict(subplot=True) def __init__(self): # Any operation involving lines objects during __init__ generates another lines object # self.candle_size = [self.data.open(i) - self.data.close(i) for i in range(0 - self.p.period + 1, 1, 1)] # returns list [-0.2,-0.1, 0.05] of len period # super(CandlePower, self).__init__() self.l.perc_of_range = bt.If(self.data.close >= self.data.open, bt.DivByZero((self.data.high - self.data.close), (self.data.high - self.data.low), zero=0.0), bt.DivByZero((self.data.close - self.data.low), (self.data.high - self.data.low), zero=0.0)) xs = np.array(list(range(1, self.p.period + 1)), dtype=np.float64) # converts [1,2,3] to array and this is a constant candle_size = self.data.close - self.data.open # lineOperation obj, even if I change this to a line, ys is still empty. ys = candle_size.get(size=self.p.period) # get the last 3 periods, returns an array. ys = np.array(ys, dtype=np.float64) # convert to np array like xs self.l.slope = ((mean(xs) * mean(ys)) - mean(xs * ys)) / ((mean(xs) * mean(xs)) - mean(xs * xs)) * 100000 # returns slope of sizes of candles to see if inc or dec def next(self): # Any operation involving lines objects during next yields regular Python types like floats and bools if 0 <= self.l.perc_of_range >= 0.25 and self.l.slope > 5: # 0 - 0.25 self.l.candle_power = 2 elif 0.76 < self.l.perc_of_range <= 1.00 and self.l.slope < - 5: # 0.76 - 1.00 self.l.candle_power = -2 else: self.l.candle_power = float('nan')
And since I am here, I like to take the chance to ask one more question.
Becasue my oop is weak, I like to ask what does super(CandlePower, self).init do? When would I need this line?
Thanks in advance!
ys = candle_size.get(size=self.p.period)
will work in the
next(), not in the