scipy 1.16.2__cp314-cp314-win_arm64.whl

scipy/interpolate/_polyint.py

@@ -0,0 +1,1025 @@
+import warnings
+
+import numpy as np
+from scipy.special import factorial
+from scipy._lib._util import (_asarray_validated, float_factorial, check_random_state,
+                              _transition_to_rng)
+
+
+__all__ = ["KroghInterpolator", "krogh_interpolate",
+           "BarycentricInterpolator", "barycentric_interpolate",
+           "approximate_taylor_polynomial"]
+
+
+def _isscalar(x):
+    """Check whether x is if a scalar type, or 0-dim"""
+    return np.isscalar(x) or hasattr(x, 'shape') and x.shape == ()
+
+
+class _Interpolator1D:
+    """
+    Common features in univariate interpolation
+
+    Deal with input data type and interpolation axis rolling. The
+    actual interpolator can assume the y-data is of shape (n, r) where
+    `n` is the number of x-points, and `r` the number of variables,
+    and use self.dtype as the y-data type.
+
+    Attributes
+    ----------
+    _y_axis
+        Axis along which the interpolation goes in the original array
+    _y_extra_shape
+        Additional trailing shape of the input arrays, excluding
+        the interpolation axis.
+    dtype
+        Dtype of the y-data arrays. Can be set via _set_dtype, which
+        forces it to be float or complex.
+
+    Methods
+    -------
+    __call__
+    _prepare_x
+    _finish_y
+    _reshape_yi
+    _set_yi
+    _set_dtype
+    _evaluate
+
+    """
+
+    __slots__ = ('_y_axis', '_y_extra_shape', 'dtype')
+
+    def __init__(self, xi=None, yi=None, axis=None):
+        self._y_axis = axis
+        self._y_extra_shape = None
+        self.dtype = None
+        if yi is not None:
+            self._set_yi(yi, xi=xi, axis=axis)
+
+    def __call__(self, x):
+        """
+        Evaluate the interpolant
+
+        Parameters
+        ----------
+        x : array_like
+            Point or points at which to evaluate the interpolant.
+
+        Returns
+        -------
+        y : array_like
+            Interpolated values. Shape is determined by replacing
+            the interpolation axis in the original array with the shape of `x`.
+
+        Notes
+        -----
+        Input values `x` must be convertible to `float` values like `int`
+        or `float`.
+
+        """
+        x, x_shape = self._prepare_x(x)
+        y = self._evaluate(x)
+        return self._finish_y(y, x_shape)
+
+    def _evaluate(self, x):
+        """
+        Actually evaluate the value of the interpolator.
+        """
+        raise NotImplementedError()
+
+    def _prepare_x(self, x):
+        """Reshape input x array to 1-D"""
+        x = _asarray_validated(x, check_finite=False, as_inexact=True)
+        x_shape = x.shape
+        return x.ravel(), x_shape
+
+    def _finish_y(self, y, x_shape):
+        """Reshape interpolated y back to an N-D array similar to initial y"""
+        y = y.reshape(x_shape + self._y_extra_shape)
+        if self._y_axis != 0 and x_shape != ():
+            nx = len(x_shape)
+            ny = len(self._y_extra_shape)
+            s = (list(range(nx, nx + self._y_axis))
+                 + list(range(nx)) + list(range(nx+self._y_axis, nx+ny)))
+            y = y.transpose(s)
+        return y
+
+    def _reshape_yi(self, yi, check=False):
+        yi = np.moveaxis(np.asarray(yi), self._y_axis, 0)
+        if check and yi.shape[1:] != self._y_extra_shape:
+            ok_shape = (f"{self._y_extra_shape[-self._y_axis:]!r} + (N,) + "
+                        f"{self._y_extra_shape[:-self._y_axis]!r}")
+            raise ValueError(f"Data must be of shape {ok_shape}")
+        return yi.reshape((yi.shape[0], -1))
+
+    def _set_yi(self, yi, xi=None, axis=None):
+        if axis is None:
+            axis = self._y_axis
+        if axis is None:
+            raise ValueError("no interpolation axis specified")
+
+        yi = np.asarray(yi)
+
+        shape = yi.shape
+        if shape == ():
+            shape = (1,)
+        if xi is not None and shape[axis] != len(xi):
+            raise ValueError("x and y arrays must be equal in length along "
+                             "interpolation axis.")
+
+        self._y_axis = (axis % yi.ndim)
+        self._y_extra_shape = yi.shape[:self._y_axis] + yi.shape[self._y_axis+1:]
+        self.dtype = None
+        self._set_dtype(yi.dtype)
+
+    def _set_dtype(self, dtype, union=False):
+        if np.issubdtype(dtype, np.complexfloating) \
+               or np.issubdtype(self.dtype, np.complexfloating):
+            self.dtype = np.complex128
+        else:
+            if not union or self.dtype != np.complex128:
+                self.dtype = np.float64
+
+
+class _Interpolator1DWithDerivatives(_Interpolator1D):
+    def derivatives(self, x, der=None):
+        """
+        Evaluate several derivatives of the polynomial at the point `x`
+
+        Produce an array of derivatives evaluated at the point `x`.
+
+        Parameters
+        ----------
+        x : array_like
+            Point or points at which to evaluate the derivatives
+        der : int or list or None, optional
+            How many derivatives to evaluate, or None for all potentially
+            nonzero derivatives (that is, a number equal to the number
+            of points), or a list of derivatives to evaluate. This number
+            includes the function value as the '0th' derivative.
+
+        Returns
+        -------
+        d : ndarray
+            Array with derivatives; ``d[j]`` contains the jth derivative.
+            Shape of ``d[j]`` is determined by replacing the interpolation
+            axis in the original array with the shape of `x`.
+
+        Examples
+        --------
+        >>> from scipy.interpolate import KroghInterpolator
+        >>> KroghInterpolator([0,0,0],[1,2,3]).derivatives(0)
+        array([1.0,2.0,3.0])
+        >>> KroghInterpolator([0,0,0],[1,2,3]).derivatives([0,0])
+        array([[1.0,1.0],
+               [2.0,2.0],
+               [3.0,3.0]])
+
+        """
+        x, x_shape = self._prepare_x(x)
+        y = self._evaluate_derivatives(x, der)
+
+        y = y.reshape((y.shape[0],) + x_shape + self._y_extra_shape)
+        if self._y_axis != 0 and x_shape != ():
+            nx = len(x_shape)
+            ny = len(self._y_extra_shape)
+            s = ([0] + list(range(nx+1, nx + self._y_axis+1))
+                 + list(range(1, nx+1)) +
+                 list(range(nx+1+self._y_axis, nx+ny+1)))
+            y = y.transpose(s)
+        return y
+
+    def derivative(self, x, der=1):
+        """
+        Evaluate a single derivative of the polynomial at the point `x`.
+
+        Parameters
+        ----------
+        x : array_like
+            Point or points at which to evaluate the derivatives
+
+        der : integer, optional
+            Which derivative to evaluate (default: first derivative).
+            This number includes the function value as 0th derivative.
+
+        Returns
+        -------
+        d : ndarray
+            Derivative interpolated at the x-points. Shape of `d` is
+            determined by replacing the interpolation axis in the
+            original array with the shape of `x`.
+
+        Notes
+        -----
+        This may be computed by evaluating all derivatives up to the desired
+        one (using self.derivatives()) and then discarding the rest.
+
+        """
+        x, x_shape = self._prepare_x(x)
+        y = self._evaluate_derivatives(x, der+1)
+        return self._finish_y(y[der], x_shape)
+
+    def _evaluate_derivatives(self, x, der=None):
+        """
+        Actually evaluate the derivatives.
+
+        Parameters
+        ----------
+        x : array_like
+            1D array of points at which to evaluate the derivatives
+        der : integer, optional
+            The number of derivatives to evaluate, from 'order 0' (der=1)
+            to order der-1.  If omitted, return all possibly-non-zero
+            derivatives, ie 0 to order n-1.
+
+        Returns
+        -------
+        d : ndarray
+            Array of shape ``(der, x.size, self.yi.shape[1])`` containing
+            the derivatives from 0 to der-1
+        """
+        raise NotImplementedError()
+
+
+class KroghInterpolator(_Interpolator1DWithDerivatives):
+    """Krogh interpolator (C∞ smooth).
+
+    The polynomial passes through all the pairs ``(xi, yi)``. One may
+    additionally specify a number of derivatives at each point `xi`;
+    this is done by repeating the value `xi` and specifying the
+    derivatives as successive `yi` values.
+
+    Allows evaluation of the polynomial and all its derivatives.
+    For reasons of numerical stability, this function does not compute
+    the coefficients of the polynomial, although they can be obtained
+    by evaluating all the derivatives.
+
+    Parameters
+    ----------
+    xi : array_like, shape (npoints, )
+        Known x-coordinates. Must be sorted in increasing order.
+    yi : array_like, shape (..., npoints, ...)
+        Known y-coordinates. When an xi occurs two or more times in
+        a row, the corresponding yi's represent derivative values. The length of `yi`
+        along the interpolation axis must be equal to the length of `xi`. Use the
+        `axis` parameter to select the correct axis.
+    axis : int, optional
+        Axis in the `yi` array corresponding to the x-coordinate values. Defaults to
+        ``axis=0``.
+
+    Notes
+    -----
+    Be aware that the algorithms implemented here are not necessarily
+    the most numerically stable known. Moreover, even in a world of
+    exact computation, unless the x coordinates are chosen very
+    carefully - Chebyshev zeros (e.g., cos(i*pi/n)) are a good choice -
+    polynomial interpolation itself is a very ill-conditioned process
+    due to the Runge phenomenon. In general, even with well-chosen
+    x values, degrees higher than about thirty cause problems with
+    numerical instability in this code.
+
+    Based on [1]_.
+
+    References
+    ----------
+    .. [1] Krogh, "Efficient Algorithms for Polynomial Interpolation
+        and Numerical Differentiation", 1970.
+
+    Examples
+    --------
+    To produce a polynomial that is zero at 0 and 1 and has
+    derivative 2 at 0, call
+
+    >>> from scipy.interpolate import KroghInterpolator
+    >>> KroghInterpolator([0,0,1],[0,2,0])
+
+    This constructs the quadratic :math:`2x^2-2x`. The derivative condition
+    is indicated by the repeated zero in the `xi` array; the corresponding
+    yi values are 0, the function value, and 2, the derivative value.
+
+    For another example, given `xi`, `yi`, and a derivative `ypi` for each
+    point, appropriate arrays can be constructed as:
+
+    >>> import numpy as np
+    >>> rng = np.random.default_rng()
+    >>> xi = np.linspace(0, 1, 5)
+    >>> yi, ypi = rng.random((2, 5))
+    >>> xi_k, yi_k = np.repeat(xi, 2), np.ravel(np.dstack((yi,ypi)))
+    >>> KroghInterpolator(xi_k, yi_k)
+
+    To produce a vector-valued polynomial, supply a higher-dimensional
+    array for `yi`:
+
+    >>> KroghInterpolator([0,1],[[2,3],[4,5]])
+
+    This constructs a linear polynomial giving (2,3) at 0 and (4,5) at 1.
+
+    """
+
+    def __init__(self, xi, yi, axis=0):
+        super().__init__(xi, yi, axis)
+
+        self.xi = np.asarray(xi)
+        self.yi = self._reshape_yi(yi)
+        self.n, self.r = self.yi.shape
+
+        if (deg := self.xi.size) > 30:
+            warnings.warn(f"{deg} degrees provided, degrees higher than about"
+                          " thirty cause problems with numerical instability "
+                          "with 'KroghInterpolator'", stacklevel=2)
+
+        c = np.zeros((self.n+1, self.r), dtype=self.dtype)
+        c[0] = self.yi[0]
+        Vk = np.zeros((self.n, self.r), dtype=self.dtype)
+        for k in range(1, self.n):
+            s = 0
+            while s <= k and xi[k-s] == xi[k]:
+                s += 1
+            s -= 1
+            Vk[0] = self.yi[k]/float_factorial(s)
+            for i in range(k-s):
+                if xi[i] == xi[k]:
+                    raise ValueError("Elements of `xi` can't be equal.")
+                if s == 0:
+                    Vk[i+1] = (c[i]-Vk[i])/(xi[i]-xi[k])
+                else:
+                    Vk[i+1] = (Vk[i+1]-Vk[i])/(xi[i]-xi[k])
+            c[k] = Vk[k-s]
+        self.c = c
+
+    def _evaluate(self, x):
+        pi = 1
+        p = np.zeros((len(x), self.r), dtype=self.dtype)
+        p += self.c[0,np.newaxis,:]
+        for k in range(1, self.n):
+            w = x - self.xi[k-1]
+            pi = w*pi
+            p += pi[:,np.newaxis] * self.c[k]
+        return p
+
+    def _evaluate_derivatives(self, x, der=None):
+        n = self.n
+        r = self.r
+
+        if der is None:
+            der = self.n
+
+        pi = np.zeros((n, len(x)))
+        w = np.zeros((n, len(x)))
+        pi[0] = 1
+        p = np.zeros((len(x), self.r), dtype=self.dtype)
+        p += self.c[0, np.newaxis, :]
+
+        for k in range(1, n):
+            w[k-1] = x - self.xi[k-1]
+            pi[k] = w[k-1] * pi[k-1]
+            p += pi[k, :, np.newaxis] * self.c[k]
+
+        cn = np.zeros((max(der, n+1), len(x), r), dtype=self.dtype)
+        cn[:n+1, :, :] += self.c[:n+1, np.newaxis, :]
+        cn[0] = p
+        for k in range(1, n):
+            for i in range(1, n-k+1):
+                pi[i] = w[k+i-1]*pi[i-1] + pi[i]
+                cn[k] = cn[k] + pi[i, :, np.newaxis]*cn[k+i]
+            cn[k] *= float_factorial(k)
+
+        cn[n, :, :] = 0
+        return cn[:der]
+
+
+def krogh_interpolate(xi, yi, x, der=0, axis=0):
+    """Convenience function for Krogh interpolation.
+
+    See `KroghInterpolator` for more details.
+
+    Parameters
+    ----------
+    xi : array_like
+        Interpolation points (known x-coordinates).
+    yi : array_like
+        Known y-coordinates, of shape ``(xi.size, R)``. Interpreted as
+        vectors of length R, or scalars if R=1.
+    x : array_like
+        Point or points at which to evaluate the derivatives.
+    der : int or list or None, optional
+        How many derivatives to evaluate, or None for all potentially
+        nonzero derivatives (that is, a number equal to the number
+        of points), or a list of derivatives to evaluate. This number
+        includes the function value as the '0th' derivative.
+    axis : int, optional
+        Axis in the `yi` array corresponding to the x-coordinate values.
+
+    Returns
+    -------
+    d : ndarray
+        If the interpolator's values are R-D then the
+        returned array will be the number of derivatives by N by R.
+        If `x` is a scalar, the middle dimension will be dropped; if
+        the `yi` are scalars then the last dimension will be dropped.
+
+    See Also
+    --------
+    KroghInterpolator : Krogh interpolator
+
+    Notes
+    -----
+    Construction of the interpolating polynomial is a relatively expensive
+    process. If you want to evaluate it repeatedly consider using the class
+    KroghInterpolator (which is what this function uses).
+
+    Examples
+    --------
+    We can interpolate 2D observed data using Krogh interpolation:
+
+    >>> import numpy as np
+    >>> import matplotlib.pyplot as plt
+    >>> from scipy.interpolate import krogh_interpolate
+    >>> x_observed = np.linspace(0.0, 10.0, 11)
+    >>> y_observed = np.sin(x_observed)
+    >>> x = np.linspace(min(x_observed), max(x_observed), num=100)
+    >>> y = krogh_interpolate(x_observed, y_observed, x)
+    >>> plt.plot(x_observed, y_observed, "o", label="observation")
+    >>> plt.plot(x, y, label="krogh interpolation")
+    >>> plt.legend()
+    >>> plt.show()
+    """
+
+    P = KroghInterpolator(xi, yi, axis=axis)
+    if der == 0:
+        return P(x)
+    elif _isscalar(der):
+        return P.derivative(x, der=der)
+    else:
+        return P.derivatives(x, der=np.amax(der)+1)[der]
+
+
+def approximate_taylor_polynomial(f,x,degree,scale,order=None):
+    """
+    Estimate the Taylor polynomial of f at x by polynomial fitting.
+
+    Parameters
+    ----------
+    f : callable
+        The function whose Taylor polynomial is sought. Should accept
+        a vector of `x` values.
+    x : scalar
+        The point at which the polynomial is to be evaluated.
+    degree : int
+        The degree of the Taylor polynomial
+    scale : scalar
+        The width of the interval to use to evaluate the Taylor polynomial.
+        Function values spread over a range this wide are used to fit the
+        polynomial. Must be chosen carefully.
+    order : int or None, optional
+        The order of the polynomial to be used in the fitting; `f` will be
+        evaluated ``order+1`` times. If None, use `degree`.
+
+    Returns
+    -------
+    p : poly1d instance
+        The Taylor polynomial (translated to the origin, so that
+        for example p(0)=f(x)).
+
+    Notes
+    -----
+    The appropriate choice of "scale" is a trade-off; too large and the
+    function differs from its Taylor polynomial too much to get a good
+    answer, too small and round-off errors overwhelm the higher-order terms.
+    The algorithm used becomes numerically unstable around order 30 even
+    under ideal circumstances.
+
+    Choosing order somewhat larger than degree may improve the higher-order
+    terms.
+
+    Examples
+    --------
+    We can calculate Taylor approximation polynomials of sin function with
+    various degrees:
+
+    >>> import numpy as np
+    >>> import matplotlib.pyplot as plt
+    >>> from scipy.interpolate import approximate_taylor_polynomial
+    >>> x = np.linspace(-10.0, 10.0, num=100)
+    >>> plt.plot(x, np.sin(x), label="sin curve")
+    >>> for degree in np.arange(1, 15, step=2):
+    ...     sin_taylor = approximate_taylor_polynomial(np.sin, 0, degree, 1,
+    ...                                                order=degree + 2)
+    ...     plt.plot(x, sin_taylor(x), label=f"degree={degree}")
+    >>> plt.legend(bbox_to_anchor=(1.05, 1), loc='upper left',
+    ...            borderaxespad=0.0, shadow=True)
+    >>> plt.tight_layout()
+    >>> plt.axis([-10, 10, -10, 10])
+    >>> plt.show()
+
+    """
+    if order is None:
+        order = degree
+
+    n = order+1
+    # Choose n points that cluster near the endpoints of the interval in
+    # a way that avoids the Runge phenomenon. Ensure, by including the
+    # endpoint or not as appropriate, that one point always falls at x
+    # exactly.
+    xs = scale*np.cos(np.linspace(0,np.pi,n,endpoint=n % 1)) + x
+
+    P = KroghInterpolator(xs, f(xs))
+    d = P.derivatives(x,der=degree+1)
+
+    return np.poly1d((d/factorial(np.arange(degree+1)))[::-1])
+
+
+class BarycentricInterpolator(_Interpolator1DWithDerivatives):
+    r"""Barycentric (Lagrange with improved stability) interpolator (C∞ smooth).
+
+    Constructs a polynomial that passes through a given set of points.
+    Allows evaluation of the polynomial and all its derivatives,
+    efficient changing of the y-values to be interpolated,
+    and updating by adding more x- and y-values. For numerical stability, a barycentric
+    representation is used rather than computing the coefficients of the polynomial
+    directly.
+
+
+    Parameters
+    ----------
+    xi : array_like, shape (npoints, )
+        1-D array of x-coordinates of the points the polynomial
+        should pass through
+    yi : array_like, shape (..., npoints, ...), optional
+        N-D array of y-coordinates of the points the polynomial should pass through.
+        If None, the y values will be supplied later via the `set_y` method.
+        The length of `yi` along the interpolation axis must be equal to the length
+        of `xi`. Use the ``axis`` parameter to select correct axis.
+    axis : int, optional
+        Axis in the yi array corresponding to the x-coordinate values. Defaults
+        to ``axis=0``.
+    wi : array_like, optional
+        The barycentric weights for the chosen interpolation points `xi`.
+        If absent or None, the weights will be computed from `xi` (default).
+        This allows for the reuse of the weights `wi` if several interpolants
+        are being calculated using the same nodes `xi`, without re-computation. This
+        also allows for computing the weights explicitly for some choices of
+        `xi` (see notes).
+    rng : {None, int, `numpy.random.Generator`}, optional
+        If `rng` is passed by keyword, types other than `numpy.random.Generator` are
+        passed to `numpy.random.default_rng` to instantiate a ``Generator``.
+        If `rng` is already a ``Generator`` instance, then the provided instance is
+        used. Specify `rng` for repeatable interpolation.
+
+        If this argument `random_state` is passed by keyword,
+        legacy behavior for the argument `random_state` applies:
+
+        - If `random_state` is None (or `numpy.random`), the `numpy.random.RandomState`
+          singleton is used.
+        - If `random_state` is an int, a new ``RandomState`` instance is used,
+          seeded with `random_state`.
+        - If `random_state` is already a ``Generator`` or ``RandomState`` instance then
+          that instance is used.
+
+        .. versionchanged:: 1.15.0
+            As part of the `SPEC-007 <https://scientific-python.org/specs/spec-0007/>`_
+            transition from use of `numpy.random.RandomState` to
+            `numpy.random.Generator` this keyword was changed from `random_state` to `rng`.
+            For an interim period, both keywords will continue to work (only specify
+            one of them). After the interim period using the `random_state` keyword will emit
+            warnings. The behavior of the `random_state` and `rng` keywords is outlined above.
+
+    Notes
+    -----
+    This method is a variant of Lagrange polynomial interpolation [1]_ based on [2]_.
+    Instead of using Lagrange's or Newton's formula, the polynomial is represented by
+    the barycentric formula
+
+    .. math::
+
+        p(x) =
+        \frac{\sum_{i=1}^m\ w_i y_i / (x - x_i)}{\sum_{i=1}^m w_i / (x - x_i)},
+
+    where :math:`w_i` are the barycentric weights computed with the general formula
+
+    .. math::
+
+        w_i = \left( \prod_{k \neq i} x_i - x_k \right)^{-1}.
+
+    This is the same barycentric form used by `AAA` and `FloaterHormannInterpolator`.
+    However, in contrast, the weights :math:`w_i` are defined such that
+    :math:`p(x)` is a polynomial rather than a rational function.
+
+    The barycentric representation avoids many of the problems associated with
+    polynomial interpolation caused by floating-point arithmetic. However, it does not
+    avoid issues that are intrinsic to polynomial interpolation. Namely, if the
+    x-coordinates are equally spaced, then the weights can be computed explicitly using
+    the formula from [2]_
+
+    .. math::
+
+        w_i = (-1)^i {n \choose i},
+
+    where :math:`n` is the number of x-coordinates. As noted in [2]_, this means that
+    for large :math:`n` the weights vary by exponentially large factors, leading to the
+    Runge phenomenon.
+
+    To avoid this ill-conditioning, the x-coordinates should be clustered at the
+    endpoints of the interval. An excellent choice of points on the interval
+    :math:`[a,b]` are Chebyshev points of the second kind
+
+    .. math::
+
+        x_i = \frac{a + b}{2} + \frac{a - b}{2}\cos(i\pi/n).
+
+    in which case the weights can be computed explicitly as
+
+    .. math::
+
+        w_i = \begin{cases}
+                  (-1)^i/2 & i = 0,n \\
+                  (-1)^i   & \text{otherwise}
+              \end{cases}.
+
+    See [2]_ for more infomation. Note that for large :math:`n`, computing the weights
+    explicitly (see examples) will be faster than the generic formula.
+
+    References
+    ----------
+    .. [1] https://en.wikipedia.org/wiki/Lagrange_polynomial
+    .. [2] Jean-Paul Berrut and Lloyd N. Trefethen, "Barycentric Lagrange
+           Interpolation", SIAM Review 2004 46:3, 501-517
+           :doi:`10.1137/S0036144502417715`
+
+    Examples
+    --------
+    To produce a quintic barycentric interpolant approximating the function
+    :math:`\sin x`, and its first four derivatives, using six randomly-spaced
+    nodes in :math:`(0, \pi/2)`:
+
+    >>> import numpy as np
+    >>> import matplotlib.pyplot as plt
+    >>> from scipy.interpolate import BarycentricInterpolator
+    >>> rng = np.random.default_rng()
+    >>> xi = rng.random(6) * np.pi/2
+    >>> f, f_d1, f_d2, f_d3, f_d4 = np.sin, np.cos, lambda x: -np.sin(x), lambda x: -np.cos(x), np.sin
+    >>> P = BarycentricInterpolator(xi, f(xi), random_state=rng)
+    >>> fig, axs = plt.subplots(5, 1, sharex=True, layout='constrained', figsize=(7,10))
+    >>> x = np.linspace(0, np.pi, 100)
+    >>> axs[0].plot(x, P(x), 'r:', x, f(x), 'k--', xi, f(xi), 'xk')
+    >>> axs[1].plot(x, P.derivative(x), 'r:', x, f_d1(x), 'k--', xi, f_d1(xi), 'xk')
+    >>> axs[2].plot(x, P.derivative(x, 2), 'r:', x, f_d2(x), 'k--', xi, f_d2(xi), 'xk')
+    >>> axs[3].plot(x, P.derivative(x, 3), 'r:', x, f_d3(x), 'k--', xi, f_d3(xi), 'xk')
+    >>> axs[4].plot(x, P.derivative(x, 4), 'r:', x, f_d4(x), 'k--', xi, f_d4(xi), 'xk')
+    >>> axs[0].set_xlim(0, np.pi)
+    >>> axs[4].set_xlabel(r"$x$")
+    >>> axs[4].set_xticks([i * np.pi / 4 for i in range(5)],
+    ...                   ["0", r"$\frac{\pi}{4}$", r"$\frac{\pi}{2}$", r"$\frac{3\pi}{4}$", r"$\pi$"])
+    >>> for ax, label in zip(axs, ("$f(x)$", "$f'(x)$", "$f''(x)$", "$f^{(3)}(x)$", "$f^{(4)}(x)$")):
+    ...     ax.set_ylabel(label)
+    >>> labels = ['Interpolation nodes', 'True function $f$', 'Barycentric interpolation']
+    >>> axs[0].legend(axs[0].get_lines()[::-1], labels, bbox_to_anchor=(0., 1.02, 1., .102),
+    ...               loc='lower left', ncols=3, mode="expand", borderaxespad=0., frameon=False)
+    >>> plt.show()
+
+    Next, we show how using Chebyshev points of the second kind avoids the avoids the
+    Runge phenomenon. In this example, we also compute the weights explicitly.
+
+    >>> n = 20
+    >>> def f(x): return np.abs(x) + 0.5*x - x**2
+    >>> i = np.arange(n)
+    >>> x_cheb = np.cos(i*np.pi/(n - 1))  # Chebyshev points on [-1, 1]
+    >>> w_i_cheb = (-1.)**i  # Explicit formula for weights of Chebyshev points
+    >>> w_i_cheb[[0, -1]] /= 2
+    >>> p_cheb = BarycentricInterpolator(x_cheb, f(x_cheb), wi=w_i_cheb)
+    >>> x_equi = np.linspace(-1, 1, n)
+    >>> p_equi = BarycentricInterpolator(x_equi, f(x_equi))
+    >>> xx = np.linspace(-1, 1, 1000)
+    >>> fig, ax = plt.subplots()
+    >>> ax.plot(xx, f(xx), label="Original Function")
+    >>> ax.plot(xx, p_cheb(xx), "--", label="Chebshev Points")
+    >>> ax.plot(xx, p_equi(xx), "--", label="Equally Spaced Points")
+    >>> ax.set(xlabel="$x$", ylabel="$f(x)$", xlim=[-1, 1])
+    >>> ax.legend()
+    >>> plt.show()
+
+    """ # numpy/numpydoc#87  # noqa: E501
+
+    @_transition_to_rng("random_state", replace_doc=False)
+    def __init__(self, xi, yi=None, axis=0, *, wi=None, rng=None):
+        super().__init__(xi, yi, axis)
+
+        rng = check_random_state(rng)
+
+        self.xi = np.asarray(xi, dtype=np.float64)
+        self.set_yi(yi)
+        self.n = len(self.xi)
+
+        # cache derivative object to avoid re-computing the weights with every call.
+        self._diff_cij = None
+
+        if wi is not None:
+            self.wi = wi
+        else:
+            # See page 510 of Berrut and Trefethen 2004 for an explanation of the
+            # capacity scaling and the suggestion of using a random permutation of
+            # the input factors.
+            # At the moment, the permutation is not performed for xi that are
+            # appended later through the add_xi interface. It's not clear to me how
+            # to implement that and it seems that most situations that require
+            # these numerical stability improvements will be able to provide all
+            # the points to the constructor.
+            self._inv_capacity = 4.0 / (np.max(self.xi) - np.min(self.xi))
+            permute = rng.permutation(self.n, )
+            inv_permute = np.zeros(self.n, dtype=np.int32)
+            inv_permute[permute] = np.arange(self.n)
+            self.wi = np.zeros(self.n)
+
+            for i in range(self.n):
+                dist = self._inv_capacity * (self.xi[i] - self.xi[permute])
+                dist[inv_permute[i]] = 1.0
+                prod = np.prod(dist)
+                if prod == 0.0:
+                    raise ValueError("Interpolation points xi must be"
+                                     " distinct.")
+                self.wi[i] = 1.0 / prod
+
+    def set_yi(self, yi, axis=None):
+        """
+        Update the y values to be interpolated
+
+        The barycentric interpolation algorithm requires the calculation
+        of weights, but these depend only on the `xi`. The `yi` can be changed
+        at any time.
+
+        Parameters
+        ----------
+        yi : array_like
+            The y-coordinates of the points the polynomial will pass through.
+            If None, the y values must be supplied later.
+        axis : int, optional
+            Axis in the `yi` array corresponding to the x-coordinate values.
+
+        """
+        if yi is None:
+            self.yi = None
+            return
+        self._set_yi(yi, xi=self.xi, axis=axis)
+        self.yi = self._reshape_yi(yi)
+        self.n, self.r = self.yi.shape
+        self._diff_baryint = None
+
+    def add_xi(self, xi, yi=None):
+        """
+        Add more x values to the set to be interpolated
+
+        The barycentric interpolation algorithm allows easy updating by
+        adding more points for the polynomial to pass through.
+
+        Parameters
+        ----------
+        xi : array_like
+            The x coordinates of the points that the polynomial should pass
+            through.
+        yi : array_like, optional
+            The y coordinates of the points the polynomial should pass through.
+            Should have shape ``(xi.size, R)``; if R > 1 then the polynomial is
+            vector-valued.
+            If `yi` is not given, the y values will be supplied later. `yi`
+            should be given if and only if the interpolator has y values
+            specified.
+
+        Notes
+        -----
+        The new points added by `add_xi` are not randomly permuted
+        so there is potential for numerical instability,
+        especially for a large number of points. If this
+        happens, please reconstruct interpolation from scratch instead.
+        """
+        if yi is not None:
+            if self.yi is None:
+                raise ValueError("No previous yi value to update!")
+            yi = self._reshape_yi(yi, check=True)
+            self.yi = np.vstack((self.yi,yi))
+        else:
+            if self.yi is not None:
+                raise ValueError("No update to yi provided!")
+        old_n = self.n
+        self.xi = np.concatenate((self.xi,xi))
+        self.n = len(self.xi)
+        self.wi **= -1
+        old_wi = self.wi
+        self.wi = np.zeros(self.n)
+        self.wi[:old_n] = old_wi
+        for j in range(old_n, self.n):
+            self.wi[:j] *= self._inv_capacity * (self.xi[j]-self.xi[:j])
+            self.wi[j] = np.multiply.reduce(
+                self._inv_capacity * (self.xi[:j]-self.xi[j])
+            )
+        self.wi **= -1
+        self._diff_cij = None
+        self._diff_baryint = None
+
+    def __call__(self, x):
+        """Evaluate the interpolating polynomial at the points x
+
+        Parameters
+        ----------
+        x : array_like
+            Point or points at which to evaluate the interpolant.
+
+        Returns
+        -------
+        y : array_like
+            Interpolated values. Shape is determined by replacing
+            the interpolation axis in the original array with the shape of `x`.
+
+        Notes
+        -----
+        Currently the code computes an outer product between `x` and the
+        weights, that is, it constructs an intermediate array of size
+        ``(N, len(x))``, where N is the degree of the polynomial.
+        """
+        return _Interpolator1D.__call__(self, x)
+
+    def _evaluate(self, x):
+        if x.size == 0:
+            p = np.zeros((0, self.r), dtype=self.dtype)
+        else:
+            c = x[..., np.newaxis] - self.xi
+            z = c == 0
+            c[z] = 1
+            c = self.wi / c
+            with np.errstate(divide='ignore'):
+                p = np.dot(c, self.yi) / np.sum(c, axis=-1)[..., np.newaxis]
+            # Now fix where x==some xi
+            r = np.nonzero(z)
+            if len(r) == 1:  # evaluation at a scalar
+                if len(r[0]) > 0:  # equals one of the points
+                    p = self.yi[r[0][0]]
+            else:
+                p[r[:-1]] = self.yi[r[-1]]
+        return p
+
+    def derivative(self, x, der=1):
+        """
+        Evaluate a single derivative of the polynomial at the point x.
+
+        Parameters
+        ----------
+        x : array_like
+            Point or points at which to evaluate the derivatives
+        der : integer, optional
+            Which derivative to evaluate (default: first derivative).
+            This number includes the function value as 0th derivative.
+
+        Returns
+        -------
+        d : ndarray
+            Derivative interpolated at the x-points. Shape of `d` is
+            determined by replacing the interpolation axis in the
+            original array with the shape of `x`.
+        """
+        x, x_shape = self._prepare_x(x)
+        y = self._evaluate_derivatives(x, der+1, all_lower=False)
+        return self._finish_y(y, x_shape)
+
+    def _evaluate_derivatives(self, x, der=None, all_lower=True):
+        # NB: der here is not the order of the highest derivative;
+        # instead, it is the size of the derivatives matrix that
+        # would be returned with all_lower=True, including the
+        # '0th' derivative (the undifferentiated function).
+        # E.g. to evaluate the 5th derivative alone, call
+        # _evaluate_derivatives(x, der=6, all_lower=False).
+
+        if (not all_lower) and (x.size == 0 or self.r == 0):
+            return np.zeros((0, self.r), dtype=self.dtype)
+
+        if (not all_lower) and der == 1:
+            return self._evaluate(x)
+
+        if (not all_lower) and (der > self.n):
+            return np.zeros((len(x), self.r), dtype=self.dtype)
+
+        if der is None:
+            der = self.n
+
+        if all_lower and (x.size == 0 or self.r == 0):
+            return np.zeros((der, len(x), self.r), dtype=self.dtype)
+
+        if self._diff_cij is None:
+            # c[i,j] = xi[i] - xi[j]
+            c = self.xi[:, np.newaxis] - self.xi
+
+            # avoid division by 0 (diagonal entries are so far zero by construction)
+            np.fill_diagonal(c, 1)
+
+            # c[i,j] = (w[j] / w[i]) / (xi[i] - xi[j]) (equation 9.4)
+            c = self.wi/ (c * self.wi[..., np.newaxis])
+
+            # fill in correct diagonal entries: each column sums to 0
+            np.fill_diagonal(c, 0)
+
+            # calculate diagonal
+            # c[j,j] = -sum_{i != j} c[i,j] (equation 9.5)
+            d = -c.sum(axis=1)
+            # c[i,j] = l_j(x_i)
+            np.fill_diagonal(c, d)
+
+            self._diff_cij = c
+
+        if self._diff_baryint is None:
+            # initialise and cache derivative interpolator and cijs;
+            # reuse weights wi (which depend only on interpolation points xi),
+            # to avoid unnecessary re-computation
+            self._diff_baryint = BarycentricInterpolator(xi=self.xi,
+                                                         yi=self._diff_cij @ self.yi,
+                                                         wi=self.wi)
+            self._diff_baryint._diff_cij = self._diff_cij
+
+        if all_lower:
+            # assemble matrix of derivatives from order 0 to order der-1,
+            # in the format required by _Interpolator1DWithDerivatives.
+            cn = np.zeros((der, len(x), self.r), dtype=self.dtype)
+            for d in range(der):
+                cn[d, :, :] = self._evaluate_derivatives(x, d+1, all_lower=False)
+            return cn
+
+        # recursively evaluate only the derivative requested
+        return self._diff_baryint._evaluate_derivatives(x, der-1, all_lower=False)
+
+
+def barycentric_interpolate(xi, yi, x, axis=0, *, der=0, rng=None):
+    """Convenience function for barycentric interpolation.
+
+    Constructs a polynomial that passes through a given set of points,
+    then evaluates the polynomial. For reasons of numerical stability,
+    this function does not compute the coefficients of the polynomial.
+
+    This function uses a "barycentric interpolation" method that treats
+    the problem as a special case of rational function interpolation.
+    This algorithm is quite stable, numerically, but even in a world of
+    exact computation, unless the `x` coordinates are chosen very
+    carefully - Chebyshev zeros (e.g., cos(i*pi/n)) are a good choice -
+    polynomial interpolation itself is a very ill-conditioned process
+    due to the Runge phenomenon.
+
+    Parameters
+    ----------
+    xi : array_like
+        1-D array of x coordinates of the points the polynomial should
+        pass through
+    yi : array_like
+        The y coordinates of the points the polynomial should pass through.
+    x : scalar or array_like
+        Point or points at which to evaluate the interpolant.
+    axis : int, optional
+        Axis in the `yi` array corresponding to the x-coordinate values.
+    der : int or list or None, optional
+        How many derivatives to evaluate, or None for all potentially
+        nonzero derivatives (that is, a number equal to the number
+        of points), or a list of derivatives to evaluate. This number
+        includes the function value as the '0th' derivative.
+    rng : `numpy.random.Generator`, optional
+        Pseudorandom number generator state. When `rng` is None, a new
+        `numpy.random.Generator` is created using entropy from the
+        operating system. Types other than `numpy.random.Generator` are
+        passed to `numpy.random.default_rng` to instantiate a ``Generator``.
+
+    Returns
+    -------
+    y : scalar or array_like
+        Interpolated values. Shape is determined by replacing
+        the interpolation axis in the original array with the shape of `x`.
+
+    See Also
+    --------
+    BarycentricInterpolator : Barycentric interpolator
+
+    Notes
+    -----
+    Construction of the interpolation weights is a relatively slow process.
+    If you want to call this many times with the same xi (but possibly
+    varying yi or x) you should use the class `BarycentricInterpolator`.
+    This is what this function uses internally.
+
+    Examples
+    --------
+    We can interpolate 2D observed data using barycentric interpolation:
+
+    >>> import numpy as np
+    >>> import matplotlib.pyplot as plt
+    >>> from scipy.interpolate import barycentric_interpolate
+    >>> x_observed = np.linspace(0.0, 10.0, 11)
+    >>> y_observed = np.sin(x_observed)
+    >>> x = np.linspace(min(x_observed), max(x_observed), num=100)
+    >>> y = barycentric_interpolate(x_observed, y_observed, x)
+    >>> plt.plot(x_observed, y_observed, "o", label="observation")
+    >>> plt.plot(x, y, label="barycentric interpolation")
+    >>> plt.legend()
+    >>> plt.show()
+
+    """
+    P = BarycentricInterpolator(xi, yi, axis=axis, rng=rng)
+    if der == 0:
+        return P(x)
+    elif _isscalar(der):
+        return P.derivative(x, der=der)
+    else:
+        return P.derivatives(x, der=np.amax(der)+1)[der]