Gym_exchange#

To use AlphaTrade, first install it using pip:

(.venv) $ pip install gym_exchange

To retrieve a list of random ingredients, you can use the gym_exchange.make() function:

For example:

>>> import gym_exchange
>>> gym_exchange.make()

data_orderbook_adapter#

1 data_pipeline.py used for defining the data path of your reading data. You should config it before runnning.

2 decoder.py is used for transferring the data into the trading signals

decoder = Decoder(**DataPipeline()()) here DataPipeline() genereate an instance. DataPipeline()() call the __call__() method. **DataPipeline()() unpack the return tuple of the __call__() method of class DataPipeline.
signals_list = decoder.process() the process()` is the main method which would be used in the class Decoder. Here decoder is an instance of Decoder.
Four major components: InsideSignal, OutsideSignal, SignalPorcessor, and DataAdjuster
- the trading signals is produce by two parts: InsideSignalEncoder and OutsideSignalEncoder
  
  the InsideSignal means the trading signals inside the price level (e.g. 10 for our example data)
  
  the OutsideSignal means the trading signals outside the price level (e.g. 10 for our example data)
  
  SignalPorcessor is a class used to ensure the trading signals is right by executing the trading signals and then get the orderbook and then compare the generated orderbook with the l2 data.
- class DataAdjuster is another important part of the decoder, which helps adjust_data_drift. It means that the l2 and l3 data may contains different info, and this class is usd for extracting the useful info which might be in the l2 data, but not l3.
  
  e.g. some order is partly cancelled outside the price level

3 encoder.py is used for transfer the trading signals into order flows, the later is machine-readable for the package orderbook/jaxob

>>> decoder = Decoder(**DataPipeline()())
>>> encoder = Encoder(decoder)
>>> Ofs = encoder()