Function _quantile

numpy/lib/_function_base_impl.py:4726–4915

Private function that doesn't support extended axis or keepdims. These methods are extended to this function using _ureduce. See nanpercentile for parameter usage. It computes the quantiles of the array for the given axis. A linear interpolation is performed based on the `method

(
    arr: "np.typing.ArrayLike",
    quantiles: np.ndarray,
    axis: int = -1,
    method: str = "linear",
    out: np.ndarray | None = None,
    weights: "np.typing.ArrayLike | None" = None,
    weak_q: bool = False,
)

def _quantile(
    arr: "np.typing.ArrayLike",
    quantiles: np.ndarray,
    axis: int = -1,
    method: str = "linear",
    out: np.ndarray | None = None,
    weights: "np.typing.ArrayLike | None" = None,
    weak_q: bool = False,
) -> np.ndarray:
    """
    Private function that doesn't support extended axis or keepdims.
    These methods are extended to this function using _ureduce.
    See nanpercentile for parameter usage.
    It computes the quantiles of the array for the given axis.
    A linear interpolation is performed based on the `method`.

    By default, the method is "linear" where alpha == beta == 1 which
    performs the 7th method of Hyndman&Fan.
    With "median_unbiased" we get alpha == beta == 1/3
    thus the 8th method of Hyndman&Fan.
    """
    # --- Setup
    arr = np.asanyarray(arr)
    values_count = arr.shape[axis]
    # The dimensions of `q` are prepended to the output shape, so we need the
    # axis being sampled from `arr` to be last.
    if axis != 0:  # But moveaxis is slow, so only call it if necessary.
        arr = np.moveaxis(arr, axis, destination=0)
    supports_nans = (
        np.issubdtype(arr.dtype, np.inexact) or arr.dtype.kind in 'Mm'
    )

    if weights is None:
        # --- Computation of indexes
        # Index where to find the value in the sorted array.
        # Virtual because it is a floating point value, not a valid index.
        # The nearest neighbours are used for interpolation
        try:
            method_props = _QuantileMethods[method]
        except KeyError:
            raise ValueError(
                f"{method!r} is not a valid method. Use one of: "
                f"{_QuantileMethods.keys()}") from None
        virtual_indexes = method_props["get_virtual_index"](values_count,
                                                            quantiles)
        virtual_indexes = np.asanyarray(virtual_indexes)

        if method_props["fix_gamma"] is None:
            supports_integers = True
        else:
            int_virtual_indices = np.issubdtype(virtual_indexes.dtype,
                                                np.integer)
            supports_integers = method == 'linear' and int_virtual_indices

        if supports_integers:
            # No interpolation needed, take the points along axis
            if supports_nans:
                # may contain nan, which would sort to the end