INNER CODE UNIT · Python

load_dataset

paperswithbacktest/pwb-toolbox · pwb_toolbox/datasets/__init__.py:642

def load_dataset(
    path,
    symbols=None,
    adjust=True,
    extend=False,
    to_usd=True,
    rate_to_price=True,
    use_hf=False
):
    split = "train"
    pwb_api_key = _get_pwb_api_key()

    # Resolve the symbol filter up front (on a copy, so the caller's list is
    # never mutated) so it can be pushed down to the per-shard parquet reads
    # instead of concatenating every shard in memory and filtering afterwards.
    symbol_filter = None
    if isinstance(symbols, list):
        symbol_filter = list(symbols)

View source record →