One article of quantitative framework backtrader: understand indicator indicators

brief introduction

The previous series of articles have covered SMA、EMA、MACD、KDJ、RSI、BOLL And other technical indicators , Not seen yet , It is recommended to follow the official account to check .

Backtrader The framework has built-in more than 100 technical analysis indicators , Packaged in backtrader.indicators In bag , It can greatly reduce the development cost , Improve the speed of strategy iteration .

Usage method

Indicators Indicators can be used in 2 A place to ： One is to use ; The other is to use in other indicators .

backtrader It is very easy to use built-in indicators , just ：

• At the strategic level __init__ Instantiate the corresponding indicator in the method

• stay next Method to use or check the corresponding index value or its derived value

Here is SMA Examples of using simple moving average indicators .

import backtrader as bt

class MyStrategy(bt.Strategy):
    params = (('period', 20),)

    def __init__(self):
        self.sma = bt.indicators.SMA(self.data, period=self.p.period)
        ...

    def next(self):
       if self.sma[0] > self.data.close[0]:
           self.buy()

It should be noted that ：

• __init__ Any index declared in the method will be in next Calculate before method call

• stay __init__ Method for lines Any operation of the object will generate other line object (python operators overloading overriding), And in the next Method generates regular python type , Such as floats or bools

• __init__ Method is faster , At the same time, it can make next The logic of the method is simpler

• __init__ Method does not support partial python The operator , Need to use bt Built in functions to handle , Such as bt.And, bt.Or, bt.All, bt.Any. In addition to these ,backtrader It also provides bt.Cmp, bt.If, bt.Max, bt.Min, bt.Sum, bt.DivByZero Such as function

backtrader The list of supported native indicators can be viewed on the official website ：Indicators - Reference - Backtrader

It should be noted that backtrader It also supports indicator aliases , Such as SMA Or you could write it as MovingAverageSimple or SimpleMovingAverage

MovingAverageSimple
Alias:

SMA, SimpleMovingAverage
Non-weighted average of the last n periods

Formula:

movav = Sum(data, period) / period

Use TA-Lib

although backtrader With many indicators , It is also easier to add new indicators . But because of TA-lib Widely used , Everyone trusts it ,backtrader It is also integrated. TA-Lib.

import backtrader as bt

class MyStrategy(bt.Strategy):
    params = (('period', 20),)

    def __init__(self):
        self.sma = bt.talib.SMA(self.data, timeperiod=self.p.period)
        ...
...

It's on it backtrader Use in talib Medium SMA Example , It can be seen that there is little difference in the use method .

adopt  print(bt.talib.SMA.__doc__)  You can view the help documentation .

>>> print(bt.talib.SMA.__doc__)
SMA([input_arrays], [timeperiod=30])

Simple Moving Average (Overlap Studies)

Inputs:
    price: (any ndarray)
Parameters:
    timeperiod: 30
Outputs:
    real

backtrader Supported by talib The index list can be viewed on the official website ：Indicators - ta-lib - Reference - Backtrader

Custom metrics

#  Inherited from bt.Indicator Or other existing indicator classes 
class DummyInd(bt.Indicator):
    #  Define held lines, Need at least 1 individual line
    lines = ('dummyline',)
    
    # params Parameters can be chosen 
    params = (('value', 5),)

    # plotinfo Optional , Used to control drawing behavior 
    plotinfo = dict(subplot=False)

    # __init__ Method or next Method required 
    def __init__(self):
        self.lines.dummyline = bt.Max(0.0, self.params.value)

This indicator in the sample code will output 0.0, Or is it self.params.value, Depending on self.params.value Is it greater than 0.0.

In addition to the __init__ Method implementation , It can also be in next Method to achieve index calculation , As shown below .

    def next(self):
        self.lines.dummyline[0] = max(0.0, self.params.value)

Except for efficiency 、 Readability , When it comes to peroid Minimum cycle ,next Methods need to be handled by yourself ,__init__ Methods do not require special treatment . So the best way is __init__ Method , If it doesn't work , Just consider in next Method . You can also use once Method runonce Mode for computational optimization .

Indicator Visualization

If the program calls cerebro.plot, that

• All declared indicators will be automatically plotted

• Operation generated lines Objects will not be drawn , Such as close_over_sma = self.data.close > self.sma

• If you want to draw the lines object , have access to LinePlotterIndicator  class ,name Parameters are held by this indicator line name

close_over_sma = self.data.close > self.sma
LinePlotterIndicator(close_over_sma, name='Close_over_SMA')

Can pass plotinfo Declaration to control the drawing of indicators .plotinfo It can be tuple、dict or OrderedDict.

class MyIndicator(bt.Indicator):

    ....
    plotinfo = dict(subplot=False)
    ....

#  You can set it separately after instantiation 
myind = MyIndicator(self.data, someparam=value)
myind.plotinfo.subplot = True

#  You can also set when instantiating 
myind = MyIndicator(self.data, someparams=value, subplot=True)

plotinfo The parameter list of is ：plot( Whether to draw , The default is True),subplot( Whether to draw in a separate window , The default is True,MA Class index this parameter is False),plotname( Index map name , Default to indicator class name ),plotabove( The plot position is above the data , The default is False),plotlinelabels, plotymargin, plotyticks,plothlines, plotyhlines, plotforce.

Conclusion &  communication

Pay attention to WeChat public number ： Zhuge said talk, Get more . At the same time, you can also get an invitation to join the quantitative investment seminar group , With many practitioners 、 Technical leaders communicate with each other 、 Compare notes , Quota co., LTD. , Don't miss .

It's not easy to write , If you think this article can help you , Help point like forward appreciation , Let the author have the motivation to insist on writing a good article .

Reference resources

• Indicators - Usage - Backtrader

• Indicators - ta-lib - Reference - Backtrader

• backtrader/indicator.py at master · mementum/backtrader · GitHub