Indicators & Machine Learning

[Sahli Simo]

Lets say that I developed an indicator to predict stock prices for the following 5 days.
I followed your tutorial and I could write a custom indicator class. But I couldnt assign the values for the coming 4 days for example. Maybe the following code will show the problem

class AlphayaPredictor(bt.Indicator):
  lines = ('lstm',)
  params = (('period', 5),)

  plotinfo = dict(subplot=False)

  # def __init__(self):
  #     self.addminperiod(self.params.period)

  def next(self):
    # datasum = math.fsum(self.data.get(size=self.params.period))
    current_date = self.datas[0].datetime
    prices = [[self.data.open[-i],
                     self.data.high[-i],
                     self.data.low[-i],
                     self.data.close[-i],
                     self.data.volume[-i]
                     ] for i in range(self.params.period-1,-1,-1)]
    pred_prices = predict_future_prices(prices)    # it returns a list of future prices

    for i in range(self.params.period):
            self.lines.lstm[i] = pred_prices[i]        # <---- here is the problem
    # self.lines.lstm[0] = pred_prices[0]          # <---- this is okay, but it's not my objective.

Could u plz tell me how to do it? I appreciate your help.

[backtrader]

You cannot put values into the future, which is what you are trying to do.

Keep the values in a ring buffer (use for example collections.deque with a maxlen during creation), push the last predicted value and pop the value from the left which is the current value.