Skip to main content

"Hello World"

class Strategy(StrategyBase):
    def __init__(self):
        # strategy attributes
        self.period = 60 * 60
        self.subscribed_books = {}
        self.options = {}

        # define your attributes here
        pass

    def on_order_state_change(self,  order):
        pass

    def trade(self, candles):
        exchange, pair, base, quote = CA.get_exchange_pair()
        close_price = candles[exchange][pair][0]['close']
        CA.log('Hello world! Current price is {}'.format(close_price))

Lets start from this Python 3 example. We first declare a class Strategy which extends from the systm built-in class StrategyBase.

In the Strategy constructor, __init__(self), we initialize period, subscribed_books, and options attributes:

  • period: How frequent your strategy recieves candle data and execute trade(). This determines the candle size i.e 1 minute, 1 hour, or 1 day. The period cannot not be changed once initialized. The smallest period is 60 second
  • subscribed_books: No need to change for single pair strategy. If developing multiple pairs strategy, set up exchange and pairs here (Not suppported yet).
  • options: No need to initialize here.

Strategy class should implement the belows 2 methods.

  • on_order_state_change(self, order): When the order status has changed, the system would pass order: Order in this callback method.
  • trade(self, candles): During each period, system would pass the latest price and volume data candles: Candle[] in this callback method.

In this example, we use a CA library method CA.get_exchange_pair() to get the name of exchange, pair, base, and quote. By using exchange and pair to get the latest close price from candles.

candles[exchange][pair] returs an candles array. The array has data sorted from the latest to the oldest with at most 500 records data. The K bar includes the below fields.

    open
    high 
    low
    close
    volume
    time

We can also use CA.log() to print out the string 'Hello world! Current price is {}'.format(close_price). After backtesting, the logged messages would be shown in the records.