PythonData as Crypto/FX data

Hi tdrxy ,

Thank you for your patience.
The Crypto (and Forex) security type is modeled differently from Equity because of the cashbook component. This security type implements the concept of base currency (the base currency of BTCUSD is BTC while USD is the quote currency) that a custom data type does not have. The Lean engine will check the SecutityType (SecurityType.Base for custom data) for operations that involve the base currency.

QuantConnect provides cryptocurrency data, free of charge, in QuantConnect Cloud, so why do you want to create custom data for it?

Best regards,
Alex

Hi, 

Thanks for getting back Alexandre Catarino. We have a specific use case where we deal with very niche pairs not available in the sources QC provides sadly. 
Is there any way to trick the engine to use our data as a crypto/fx sec type?

Cheers!

Hi Tdrxy

One hack would be using a SymbolObject to reference the “currency in hand”:

def Initialize(self):
    self.symbol_data = {}
    
    self.symbol = self.AddData(CustomDataClass, “BTCUSD”).Symbol
    self.symbol_data[self.symbol] = SymbolData(self, self.symbol, ”BTC”)
    
def OnData(self, data):
    total_btc_in_hand = self.symbol_data[self.symbol].TotalQuantity

class SymbolData:
    def __init__(self, algorithm, symbol, quote_currency):
        self._algorithm = algorithm
        self.Symbol = symbol
        self.QuoteCurrency = quote_currency
    
    @property
    def TotalQuantity(self):
        portfolio = self._algorithm.Portfolio
        return portfolio[self.Symbol].Quantity + portfolio.Cashbook[self.QuoteCurrency].Amount

Best
Louis