microsoft/qlib · error · ValueError

sampling freq must be xmin, xd, xw, xm

Error message

sampling freq must be xmin, xd, xw, xm

What it means

Error "sampling freq must be xmin, xd, xw, xm" thrown in microsoft/qlib.

Source

Thrown at qlib/utils/resam.py:69

        return _calendar_minute

    # else, convert the raw calendar into day calendar, and divide the whole calendar into several bars evenly
    else:
        _calendar_day = np.unique(list(map(lambda x: pd.Timestamp(x.year, x.month, x.day, 0, 0, 0), calendar_raw)))
        if freq_sam.base == Freq.NORM_FREQ_DAY:
            return _calendar_day[:: freq_sam.count]

        elif freq_sam.base == Freq.NORM_FREQ_WEEK:
            _day_in_week = np.array(list(map(lambda x: x.dayofweek, _calendar_day)))
            _calendar_week = _calendar_day[np.ediff1d(_day_in_week, to_begin=-1) < 0]
            return _calendar_week[:: freq_sam.count]

        elif freq_sam.base == Freq.NORM_FREQ_MONTH:
            _day_in_month = np.array(list(map(lambda x: x.day, _calendar_day)))
            _calendar_month = _calendar_day[np.ediff1d(_day_in_month, to_begin=-1) < 0]
            return _calendar_month[:: freq_sam.count]
        else:
            raise ValueError("sampling freq must be xmin, xd, xw, xm")


def get_higher_eq_freq_feature(instruments, fields, start_time=None, end_time=None, freq="day", disk_cache=1):
    """get the feature with higher or equal frequency than `freq`.
    Returns
    -------
    pd.DataFrame
        the feature with higher or equal frequency
    """

    from ..data.data import D  # pylint: disable=C0415

    try:
        _result = D.features(instruments, fields, start_time, end_time, freq=freq, disk_cache=disk_cache)
        _freq = freq
    except (ValueError, KeyError) as value_key_e:
        _, norm_freq = Freq.parse(freq)
        if norm_freq in [Freq.NORM_FREQ_MONTH, Freq.NORM_FREQ_WEEK, Freq.NORM_FREQ_DAY]:

View on GitHub (pinned to 79633dd950)

Solutions

1. Use a sampling freq of the form xmin, xd, xw, or xm (e.g. '5min', '1d', '2w', '1m').
2. Correct the freq string in your call to match the supported pattern.

When it happens

Trigger: This error is raised at qlib/utils/resam.py:69 when the guard condition fails: "sampling freq must be xmin, xd, xw, xm". It triggers when the code path receives input or a state that violates the precondition checked at this point — e.g. an unimplemented abstract method being called, an unsupported argument type/value, or an invalid configuration. To avoid it, satisfy the documented precondition before this call: implement the required method in your subclass, or pass a supported value of the expected type as described by the message.

Common situations: See trigger scenarios.

AI-assisted analysis of microsoft/qlib@79633dd950 (2026-08-15). Data as JSON: /api/errors/7fb90f66d43d6bf3. Report an issue: GitHub.