schematics.yaml

Preprocess:
    - Load custom data
    Training mode:
        1. Split timeframes:
            * WINDOW: the rolling number of candles worth of data to process
            * DFS: 4 timeframes of data within the same timerange
            * LIST DATES (LD): A list of all of the dates of the timeframe used for Prediction
            - For each INDEX in LD:
                - For each DF in DFS -> Series list[4]:
                    - Create data slice of size WINDOW -> DS
                    - Grab the last 20 candles of the DS -> Series
                    - Add that to a list -> Series list
                - Create a decision -> 'LONG' or 'SHORT':
                    - If the close price of the biggest timeframe > close price of the smallest timeframe:
                        - LONG
                    - If the close price of the biggest timeframe < the close price of the smallest timeframe:
                        - SHORT