microsoft/qlib · error · NotImplementedError

This type of input is not supported

Error message

This type of input is not supported

What it means

Raised by SepDFLocGetter.__getitem__ (the object behind SepDataFrame.loc(axis=1)[...]) when selecting columns and the argument is neither a str nor a tuple/list of keys. SepDataFrame stores separate per-group frames, so fancy indexing (slices, boolean arrays, callables) on axis 1 has no implementation.

Source

Thrown at qlib/contrib/data/utils/sepdf.py:168

    def __init__(self, sdf: SepDataFrame, join):
        self._sdf = sdf
        self.axis = None
        self.join = join

    def __call__(self, axis):
        self.axis = axis
        return self

    def __getitem__(self, args):
        if self.axis == 1:
            if isinstance(args, str):
                return self._sdf[args]
            elif isinstance(args, (tuple, list)):
                new_df_dict = {k: self._sdf[k] for k in args}
                return SepDataFrame(new_df_dict, join=self.join if self.join in args else args[0], skip_align=True)
            else:
                raise NotImplementedError(f"This type of input is not supported")
        elif self.axis == 0:
            return SepDataFrame(
                {k: df.loc(axis=0)[args] for k, df in self._sdf._df_dict.items()}, join=self.join, skip_align=True
            )
        else:
            df = self._sdf
            if isinstance(args, tuple):
                ax0, *ax1 = args
                if len(ax1) == 0:
                    ax1 = None
                if ax1 is not None:
                    df = df.loc(axis=1)[ax1]
                if ax0 is not None:
                    df = df.loc(axis=0)[ax0]
                return df
            else:
                return df.loc(axis=0)[args]

View on GitHub (pinned to 79633dd950)

Solutions

1. Use a string or list of column names: sdf.loc(axis=1)['KMID'] or sdf.loc(axis=1)[['KMID', 'KLEN']].
2. Convert numpy arrays to lists: sdf.loc(axis=1)[list(col_array)].
3. For slice-based selection, first materialize the underlying pandas frame via sdf._df_dict[sdf.join] and use plain pandas .loc.

Example fix

// before
sub = sdf.loc(axis=1)[np.array(["KMID", "KLEN"])]  # NotImplementedError

// after
sub = sdf.loc(axis=1)[["KMID", "KLEN"]]

cols = ["KMID", "KLEN"]
assert all(isinstance(c, str) for c in cols)
sub = sdf.loc(axis=1)[cols]  # list/tuple/str only

def valid_axis1_args(args) -> bool:
    return isinstance(args, str) or (isinstance(args, (tuple, list)) and all(isinstance(a, str) for a in args))

try:
    sub = sdf.loc(axis=1)[args]
except NotImplementedError:
    sub = sdf._df_dict[sdf.join].loc(axis=1)[args]  # real pandas supports slices/masks

When it happens

Trigger: Calling sdf.loc(axis=1)[args] where args is a slice (e.g. [:, 'KMID']), a boolean mask, an int, or a numpy array — only a column name or tuple/list of column names are handled at axis 1.

Common situations: Copy-pasting pandas .loc idioms (slices, boolean column masks) onto high-frequency data wrapped in SepDataFrame; passing a numpy array of column names instead of a plain list.

Related errors

• Please implement the `droplevel` method
• nfs-common is not found, please install it by execute: sudo
• Mount failed: requires sudo or permission denied
• mount {provider_uri} on {mount_path} error! Command error
• Mount failed: {e.stderr}

AI-assisted analysis of microsoft/qlib@79633dd950 (2026-08-15). Data as JSON: /api/errors/089cef04dd3e5131. Report an issue: GitHub.