Best posts made by stochastick

Maybe try something list this.. Initialize empty list during init, and add to it during next(). Finally, during stop(), combine into a df and export/use for your own purposes.

@Ronan-Dunham said in How to get the data as array of the sma1 and sma2:

class SmaCross(bt.SignalStrategy):
    def __init__(self):
        sma1, sma2 = bt.ind.SMA(period=10), bt.ind.SMA(period=30)
        crossover = bt.ind.CrossOver(sma1, sma2)
        self.signal_add(bt.SIGNAL_LONG, crossover)
        self.ma1 = []     # Initialize empty lists to hold indicator values
        self.ma2 = []     # Initialize empty lists to hold indicator values

    def next(self):
        # Create data frame columns here by appending data to a list
        self.ma1.append(sma1)   
        self.ma2.append(sma2)   

    def stop(self):
        # Combine and export pandas DataFrame here   

cerebro = bt.Cerebro()
cerebro.addstrategy(SmaCross)
store = alpaca_backtrader_api.AlpacaStore(key_id=ALPACA_API_KEY,secret_key=ALPACA_SECRET_KEY,paper=ALPACA_PAPER)
DataFactory = store.getdata

data0 = DataFactory(
        dataname=symbol_to_check,
        timeframe=bt.TimeFrame.TFrame("Minutes"),
        fromdate=pd.Timestamp('2019-11-15'),compression =1,
#         todate=pd.Timestamp('2018-11-17'),
        historical=True)

cerebro.adddata(data0)

cerebro.run()
cerebro.plot()

There is most certainly a more efficient way to accomplish what you seek, but nonetheless here is a working sample that accomplishes I believe what you are searching for. The data frame is simply saved to CSV here, but you could do whatever you want with it. Data is loaded from my own sources, but just substitute with yours.

import backtrader as bt
import pandas as pd


class SmaCross(bt.SignalStrategy):

    # Global Parameters
    globalParams = {'symbols': ['EUR_USD'],
                    'startDate': '01-01-2018',
                    'endDate': '01-05-2018'}

    def __init__(self):
        self.sma1 = bt.ind.SMA(period=10)
        self.sma2 = bt.ind.SMA(period=30)
        self.crossover = bt.ind.CrossOver(self.sma1, self.sma2)
        self.signal_add(bt.SIGNAL_LONG, self.crossover)

        # Initialize empty lists to hold values
        self.date_time = []
        self.close_px = []
        self.ma1 = []
        self.ma2 = []

    def log(self, txt, dt=None):
        """ Logging function """
        dt = dt or self.datas[0].datetime.datetime()
        print('%s, %s' % (dt.isoformat(), txt))

    def next(self):
        """Log data to console and save to lists"""
        
        self.log('%s '
                 '| O %.5f '
                 '| H %.5f '
                 '| L %.5f '
                 '| C %.5f '
                 '| SMA1 %.5f '
                 '| SMA2 %.5f '
                 %
              (self.datas[0]._name,
               self.datas[0].open[0],
               self.datas[0].high[0],
               self.datas[0].low[0],
               self.datas[0].close[0],
               self.sma1[0],
               self.sma2[0]))

        self.date_time.append(self.datas[0].datetime.datetime())
        self.close_px.append(self.datas[0].close[0])
        self.ma1.append(self.sma1[0])
        self.ma2.append(self.sma2[0])

    def stop(self):
        df = pd.DataFrame(index=self.date_time)
        df['Close'] = self.close_px
        df['SMA1'] = self.ma1
        df['SMA2'] = self.ma2
        df.to_csv('trade_history.csv', sep='\t')


cerebro = bt.Cerebro()
cerebro.addstrategy(SmaCross)

symbols = SmaCross.globalParams['symbols']

for symbol in symbols:
    mktData = pd.read_json('../sample-data/FX/training_' + symbol + '.json')
    mktData['time'] = pd.to_datetime(mktData['time'])
    timeSlice = ((mktData['time'] > SmaCross.globalParams['startDate']) &
                 (mktData['time'] < SmaCross.globalParams['endDate']))
    mktData = mktData.loc[timeSlice]
    mktData = bt.feeds.PandasData(dataname=mktData, timeframe=bt.TimeFrame.Minutes, datetime='time', close='close')
    cerebro.resampledata(mktData, timeframe=bt.TimeFrame.Minutes, compression=5, name=symbol)

cerebro.run()

This creates a CSV file formatted as such.