tensortrade.exchanges.simulated.gan_exchange module

class tensortrade.exchanges.simulated.gan_exchange.GANExchange(training_data=None, **kwargs)[source]

Bases: tensortrade.exchanges.instrument_exchange.InstrumentExchange

A simulated instrument exchange, in which the price history is based off a generative adversarial network model with supplied parameters.

If the training_data parameter is not supplied upon initialization, it must be set before the exchange can be used within a trading environment.

reset()[source]

Reset the feature pipeline, initial balance, trades, performance, and any other temporary stateful data.