microsoft/qlib · critical · ValueError

provider_uri cannot be None

Error message

provider_uri cannot be None

What it means

QlibConfig.DataPathManager.format_provider_uri in qlib/config.py normalizes the provider_uri argument into a {freq: uri} dict. None is explicitly rejected with ValueError because data access requires at least one concrete location. This typically surfaces when qlib.init is called with provider_uri=None (the default of a partially-filled config) or a nested call re-formats the uri.

Source

Thrown at qlib/config.py:364

        """
        Motivation:
        - get the right path (e.g. data uri) for accessing data based on given information(e.g. provider_uri, mount_path and frequency)
        - some helper functions to process uri.
        """

        def __init__(self, provider_uri: Union[str, Path, dict], mount_path: Union[str, Path, dict]):
            """
            The relation of `provider_uri` and `mount_path`
            - `mount_path` is used only if provider_uri is an NFS path
            - otherwise, provider_uri will be used for accessing data
            """
            self.provider_uri = provider_uri
            self.mount_path = mount_path

        @staticmethod
        def format_provider_uri(provider_uri: Union[str, dict, Path]) -> dict:
            if provider_uri is None:
                raise ValueError("provider_uri cannot be None")
            if isinstance(provider_uri, (str, dict, Path)):
                if not isinstance(provider_uri, dict):
                    provider_uri = {QlibConfig.DEFAULT_FREQ: provider_uri}
            else:
                raise TypeError(f"provider_uri does not support {type(provider_uri)}")
            for freq, _uri in provider_uri.items():
                if QlibConfig.DataPathManager.get_uri_type(_uri) == QlibConfig.LOCAL_URI:
                    provider_uri[freq] = str(Path(_uri).expanduser().resolve())
            return provider_uri

        @staticmethod
        def get_uri_type(uri: Union[str, Path]):
            uri = uri if isinstance(uri, str) else str(uri.expanduser().resolve())
            is_win = re.match("^[a-zA-Z]:.*", uri) is not None  # such as 'C:\\data', 'D:'
            # such as 'host:/data/'   (User may define short hostname by themselves or use localhost)
            is_nfs_or_win = re.match("^[^/]+:.+", uri) is not None

            if is_nfs_or_win and not is_win:

View on GitHub (pinned to 79633dd950)

Solutions

  1. Set a concrete provider_uri in qlib.init (e.g. '~/.qlib/qlib_data/cn_data').
  2. Guard before init: if provider_uri from your own config is None, either fail fast with a clear message or substitute the default path.
  3. For multi-frequency data pass a dict like {'day': uri_day, '1min': uri_min} — never None values.

Example fix

# before
qlib.init(provider_uri=cfg.get('provider_uri'))  # None if key absent
# after
qlib.init(provider_uri=cfg.get('provider_uri', '~/.qlib/qlib_data/cn_data'))
Defensive patterns

Strategy: validation

Validate before calling

if provider_uri is None:
    raise ValueError('provider_uri is required — set it to a data path such as ~/.qlib/qlib_data/cn_data')

Type guard

def is_valid_provider_uri(u) -> bool:
    return u is not None

Prevention

When it happens

Trigger: qlib.init(provider_uri=None) or omitting provider_uri while some code path still constructs DataPathManager; building QlibConfig programmatically and passing a None obtained from config.get('provider_uri').

Common situations: Config templating where provider_uri is conditionally set and resolves to None; refactoring that reads the key with .get() before validating presence; upgrading qlib versions where None was previously tolerated until data load time.

Related errors


AI-assisted analysis of microsoft/qlib@79633dd950 (2026-08-15). Data as JSON: /api/errors/605078211a753bc1. Report an issue: GitHub.