first commit

dashboard/flask-server/venv/Lib/site-packages/sklearn/_loss/__init__.py

@@ -0,0 +1,29 @@
+"""
+The :mod:`sklearn._loss` module includes loss function classes suitable for
+fitting classification and regression tasks.
+"""
+
+from .loss import (
+    HalfSquaredError,
+    AbsoluteError,
+    PinballLoss,
+    HalfPoissonLoss,
+    HalfGammaLoss,
+    HalfTweedieLoss,
+    HalfTweedieLossIdentity,
+    HalfBinomialLoss,
+    HalfMultinomialLoss,
+)
+
+
+__all__ = [
+    "HalfSquaredError",
+    "AbsoluteError",
+    "PinballLoss",
+    "HalfPoissonLoss",
+    "HalfGammaLoss",
+    "HalfTweedieLoss",
+    "HalfTweedieLossIdentity",
+    "HalfBinomialLoss",
+    "HalfMultinomialLoss",
+]

dashboard/flask-server/venv/Lib/site-packages/sklearn/_loss/_loss.pxd

@@ -0,0 +1,82 @@
+# cython: language_level=3
+
+import numpy as np
+cimport numpy as np
+
+np.import_array()
+
+
+# Fused types for y_true, y_pred, raw_prediction
+ctypedef fused Y_DTYPE_C:
+    np.npy_float64
+    np.npy_float32
+
+
+# Fused types for gradient and hessian
+ctypedef fused G_DTYPE_C:
+    np.npy_float64
+    np.npy_float32
+
+
+# Struct to return 2 doubles
+ctypedef struct double_pair:
+   double val1
+   double val2
+
+
+# C base class for loss functions
+cdef class CyLossFunction:
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyHalfSquaredError(CyLossFunction):
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyAbsoluteError(CyLossFunction):
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyPinballLoss(CyLossFunction):
+    cdef readonly double quantile  # readonly makes it accessible from Python
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyHalfPoissonLoss(CyLossFunction):
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyHalfGammaLoss(CyLossFunction):
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyHalfTweedieLoss(CyLossFunction):
+    cdef readonly double power  # readonly makes it accessible from Python
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyHalfTweedieLossIdentity(CyLossFunction):
+    cdef readonly double power  # readonly makes it accessible from Python
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil
+
+
+cdef class CyHalfBinomialLoss(CyLossFunction):
+    cdef double cy_loss(self, double y_true, double raw_prediction) nogil
+    cdef double cy_gradient(self, double y_true, double raw_prediction) nogil
+    cdef double_pair cy_grad_hess(self, double y_true, double raw_prediction) nogil

dashboard/flask-server/venv/Lib/site-packages/sklearn/_loss/glm_distribution.py

@@ -0,0 +1,373 @@
+"""
+Distribution functions used in GLM
+"""
+
+# Author: Christian Lorentzen <lorentzen.ch@googlemail.com>
+# License: BSD 3 clause
+#
+# TODO(1.3): remove file
+#       This is only used for backward compatibility in _GeneralizedLinearRegressor
+#       for the deprecated family attribute.
+
+from abc import ABCMeta, abstractmethod
+from collections import namedtuple
+import numbers
+
+import numpy as np
+from scipy.special import xlogy
+
+
+DistributionBoundary = namedtuple("DistributionBoundary", ("value", "inclusive"))
+
+
+class ExponentialDispersionModel(metaclass=ABCMeta):
+    r"""Base class for reproductive Exponential Dispersion Models (EDM).
+
+    The pdf of :math:`Y\sim \mathrm{EDM}(y_\textrm{pred}, \phi)` is given by
+
+    .. math:: p(y| \theta, \phi) = c(y, \phi)
+        \exp\left(\frac{\theta y-A(\theta)}{\phi}\right)
+        = \tilde{c}(y, \phi)
+            \exp\left(-\frac{d(y, y_\textrm{pred})}{2\phi}\right)
+
+    with mean :math:`\mathrm{E}[Y] = A'(\theta) = y_\textrm{pred}`,
+    variance :math:`\mathrm{Var}[Y] = \phi \cdot v(y_\textrm{pred})`,
+    unit variance :math:`v(y_\textrm{pred})` and
+    unit deviance :math:`d(y,y_\textrm{pred})`.
+
+    Methods
+    -------
+    deviance
+    deviance_derivative
+    in_y_range
+    unit_deviance
+    unit_deviance_derivative
+    unit_variance
+
+    References
+    ----------
+    https://en.wikipedia.org/wiki/Exponential_dispersion_model.
+    """
+
+    def in_y_range(self, y):
+        """Returns ``True`` if y is in the valid range of Y~EDM.
+
+        Parameters
+        ----------
+        y : array of shape (n_samples,)
+            Target values.
+        """
+        # Note that currently supported distributions have +inf upper bound
+
+        if not isinstance(self._lower_bound, DistributionBoundary):
+            raise TypeError(
+                "_lower_bound attribute must be of type DistributionBoundary"
+            )
+
+        if self._lower_bound.inclusive:
+            return np.greater_equal(y, self._lower_bound.value)
+        else:
+            return np.greater(y, self._lower_bound.value)
+
+    @abstractmethod
+    def unit_variance(self, y_pred):
+        r"""Compute the unit variance function.
+
+        The unit variance :math:`v(y_\textrm{pred})` determines the variance as
+        a function of the mean :math:`y_\textrm{pred}` by
+        :math:`\mathrm{Var}[Y_i] = \phi/s_i*v(y_\textrm{pred}_i)`.
+        It can also be derived from the unit deviance
+        :math:`d(y,y_\textrm{pred})` as
+
+        .. math:: v(y_\textrm{pred}) = \frac{2}{
+            \frac{\partial^2 d(y,y_\textrm{pred})}{
+            \partialy_\textrm{pred}^2}}\big|_{y=y_\textrm{pred}}
+
+        See also :func:`variance`.
+
+        Parameters
+        ----------
+        y_pred : array of shape (n_samples,)
+            Predicted mean.
+        """
+
+    @abstractmethod
+    def unit_deviance(self, y, y_pred, check_input=False):
+        r"""Compute the unit deviance.
+
+        The unit_deviance :math:`d(y,y_\textrm{pred})` can be defined by the
+        log-likelihood as
+        :math:`d(y,y_\textrm{pred}) = -2\phi\cdot
+        \left(loglike(y,y_\textrm{pred},\phi) - loglike(y,y,\phi)\right).`
+
+        Parameters
+        ----------
+        y : array of shape (n_samples,)
+            Target values.
+
+        y_pred : array of shape (n_samples,)
+            Predicted mean.
+
+        check_input : bool, default=False
+            If True raise an exception on invalid y or y_pred values, otherwise
+            they will be propagated as NaN.
+        Returns
+        -------
+        deviance: array of shape (n_samples,)
+            Computed deviance
+        """
+
+    def unit_deviance_derivative(self, y, y_pred):
+        r"""Compute the derivative of the unit deviance w.r.t. y_pred.
+
+        The derivative of the unit deviance is given by
+        :math:`\frac{\partial}{\partialy_\textrm{pred}}d(y,y_\textrm{pred})
+             = -2\frac{y-y_\textrm{pred}}{v(y_\textrm{pred})}`
+        with unit variance :math:`v(y_\textrm{pred})`.
+
+        Parameters
+        ----------
+        y : array of shape (n_samples,)
+            Target values.
+
+        y_pred : array of shape (n_samples,)
+            Predicted mean.
+        """
+        return -2 * (y - y_pred) / self.unit_variance(y_pred)
+
+    def deviance(self, y, y_pred, weights=1):
+        r"""Compute the deviance.
+
+        The deviance is a weighted sum of the per sample unit deviances,
+        :math:`D = \sum_i s_i \cdot d(y_i, y_\textrm{pred}_i)`
+        with weights :math:`s_i` and unit deviance
+        :math:`d(y,y_\textrm{pred})`.
+        In terms of the log-likelihood it is :math:`D = -2\phi\cdot
+        \left(loglike(y,y_\textrm{pred},\frac{phi}{s})
+        - loglike(y,y,\frac{phi}{s})\right)`.
+
+        Parameters
+        ----------
+        y : array of shape (n_samples,)
+            Target values.
+
+        y_pred : array of shape (n_samples,)
+            Predicted mean.
+
+        weights : {int, array of shape (n_samples,)}, default=1
+            Weights or exposure to which variance is inverse proportional.
+        """
+        return np.sum(weights * self.unit_deviance(y, y_pred))
+
+    def deviance_derivative(self, y, y_pred, weights=1):
+        r"""Compute the derivative of the deviance w.r.t. y_pred.
+
+        It gives :math:`\frac{\partial}{\partial y_\textrm{pred}}
+        D(y, \y_\textrm{pred}; weights)`.
+
+        Parameters
+        ----------
+        y : array, shape (n_samples,)
+            Target values.
+
+        y_pred : array, shape (n_samples,)
+            Predicted mean.
+
+        weights : {int, array of shape (n_samples,)}, default=1
+            Weights or exposure to which variance is inverse proportional.
+        """
+        return weights * self.unit_deviance_derivative(y, y_pred)
+
+
+class TweedieDistribution(ExponentialDispersionModel):
+    r"""A class for the Tweedie distribution.
+
+    A Tweedie distribution with mean :math:`y_\textrm{pred}=\mathrm{E}[Y]`
+    is uniquely defined by it's mean-variance relationship
+    :math:`\mathrm{Var}[Y] \propto y_\textrm{pred}^power`.
+
+    Special cases are:
+
+    ===== ================
+    Power Distribution
+    ===== ================
+    0     Normal
+    1     Poisson
+    (1,2) Compound Poisson
+    2     Gamma
+    3     Inverse Gaussian
+
+    Parameters
+    ----------
+    power : float, default=0
+            The variance power of the `unit_variance`
+            :math:`v(y_\textrm{pred}) = y_\textrm{pred}^{power}`.
+            For ``0<power<1``, no distribution exists.
+    """
+
+    def __init__(self, power=0):
+        self.power = power
+
+    @property
+    def power(self):
+        return self._power
+
+    @power.setter
+    def power(self, power):
+        # We use a property with a setter, to update lower and
+        # upper bound when the power parameter is updated e.g. in grid
+        # search.
+        if not isinstance(power, numbers.Real):
+            raise TypeError("power must be a real number, input was {0}".format(power))
+
+        if power <= 0:
+            # Extreme Stable or Normal distribution
+            self._lower_bound = DistributionBoundary(-np.Inf, inclusive=False)
+        elif 0 < power < 1:
+            raise ValueError(
+                "Tweedie distribution is only defined for power<=0 and power>=1."
+            )
+        elif 1 <= power < 2:
+            # Poisson or Compound Poisson distribution
+            self._lower_bound = DistributionBoundary(0, inclusive=True)
+        elif power >= 2:
+            # Gamma, Positive Stable, Inverse Gaussian distributions
+            self._lower_bound = DistributionBoundary(0, inclusive=False)
+        else:  # pragma: no cover
+            # this branch should be unreachable.
+            raise ValueError
+
+        self._power = power
+
+    def unit_variance(self, y_pred):
+        """Compute the unit variance of a Tweedie distribution
+        v(y_\textrm{pred})=y_\textrm{pred}**power.
+
+        Parameters
+        ----------
+        y_pred : array of shape (n_samples,)
+            Predicted mean.
+        """
+        return np.power(y_pred, self.power)
+
+    def unit_deviance(self, y, y_pred, check_input=False):
+        r"""Compute the unit deviance.
+
+        The unit_deviance :math:`d(y,y_\textrm{pred})` can be defined by the
+        log-likelihood as
+        :math:`d(y,y_\textrm{pred}) = -2\phi\cdot
+        \left(loglike(y,y_\textrm{pred},\phi) - loglike(y,y,\phi)\right).`
+
+        Parameters
+        ----------
+        y : array of shape (n_samples,)
+            Target values.
+
+        y_pred : array of shape (n_samples,)
+            Predicted mean.
+
+        check_input : bool, default=False
+            If True raise an exception on invalid y or y_pred values, otherwise
+            they will be propagated as NaN.
+        Returns
+        -------
+        deviance: array of shape (n_samples,)
+            Computed deviance
+        """
+        p = self.power
+
+        if check_input:
+            message = (
+                "Mean Tweedie deviance error with power={} can only be used on ".format(
+                    p
+                )
+            )
+            if p < 0:
+                # 'Extreme stable', y any real number, y_pred > 0
+                if (y_pred <= 0).any():
+                    raise ValueError(message + "strictly positive y_pred.")
+            elif p == 0:
+                # Normal, y and y_pred can be any real number
+                pass
+            elif 0 < p < 1:
+                raise ValueError(
+                    "Tweedie deviance is only defined for power<=0 and power>=1."
+                )
+            elif 1 <= p < 2:
+                # Poisson and compound Poisson distribution, y >= 0, y_pred > 0
+                if (y < 0).any() or (y_pred <= 0).any():
+                    raise ValueError(
+                        message + "non-negative y and strictly positive y_pred."
+                    )
+            elif p >= 2:
+                # Gamma and Extreme stable distribution, y and y_pred > 0
+                if (y <= 0).any() or (y_pred <= 0).any():
+                    raise ValueError(message + "strictly positive y and y_pred.")
+            else:  # pragma: nocover
+                # Unreachable statement
+                raise ValueError
+
+        if p < 0:
+            # 'Extreme stable', y any real number, y_pred > 0
+            dev = 2 * (
+                np.power(np.maximum(y, 0), 2 - p) / ((1 - p) * (2 - p))
+                - y * np.power(y_pred, 1 - p) / (1 - p)
+                + np.power(y_pred, 2 - p) / (2 - p)
+            )
+
+        elif p == 0:
+            # Normal distribution, y and y_pred any real number
+            dev = (y - y_pred) ** 2
+        elif p < 1:
+            raise ValueError(
+                "Tweedie deviance is only defined for power<=0 and power>=1."
+            )
+        elif p == 1:
+            # Poisson distribution
+            dev = 2 * (xlogy(y, y / y_pred) - y + y_pred)
+        elif p == 2:
+            # Gamma distribution
+            dev = 2 * (np.log(y_pred / y) + y / y_pred - 1)
+        else:
+            dev = 2 * (
+                np.power(y, 2 - p) / ((1 - p) * (2 - p))
+                - y * np.power(y_pred, 1 - p) / (1 - p)
+                + np.power(y_pred, 2 - p) / (2 - p)
+            )
+        return dev
+
+
+class NormalDistribution(TweedieDistribution):
+    """Class for the Normal (aka Gaussian) distribution."""
+
+    def __init__(self):
+        super().__init__(power=0)
+
+
+class PoissonDistribution(TweedieDistribution):
+    """Class for the scaled Poisson distribution."""
+
+    def __init__(self):
+        super().__init__(power=1)
+
+
+class GammaDistribution(TweedieDistribution):
+    """Class for the Gamma distribution."""
+
+    def __init__(self):
+        super().__init__(power=2)
+
+
+class InverseGaussianDistribution(TweedieDistribution):
+    """Class for the scaled InverseGaussianDistribution distribution."""
+
+    def __init__(self):
+        super().__init__(power=3)
+
+
+EDM_DISTRIBUTIONS = {
+    "normal": NormalDistribution,
+    "poisson": PoissonDistribution,
+    "gamma": GammaDistribution,
+    "inverse-gaussian": InverseGaussianDistribution,
+}

dashboard/flask-server/venv/Lib/site-packages/sklearn/_loss/link.py

@@ -0,0 +1,261 @@
+"""
+Module contains classes for invertible (and differentiable) link functions.
+"""
+# Author: Christian Lorentzen <lorentzen.ch@gmail.com>
+
+from abc import ABC, abstractmethod
+from dataclasses import dataclass
+
+import numpy as np
+from scipy.special import expit, logit
+from scipy.stats import gmean
+from ..utils.extmath import softmax
+
+
+@dataclass
+class Interval:
+    low: float
+    high: float
+    low_inclusive: bool
+    high_inclusive: bool
+
+    def __post_init__(self):
+        """Check that low <= high"""
+        if self.low > self.high:
+            raise ValueError(
+                f"One must have low <= high; got low={self.low}, high={self.high}."
+            )
+
+    def includes(self, x):
+        """Test whether all values of x are in interval range.
+
+        Parameters
+        ----------
+        x : ndarray
+            Array whose elements are tested to be in interval range.
+
+        Returns
+        -------
+        result : bool
+        """
+        if self.low_inclusive:
+            low = np.greater_equal(x, self.low)
+        else:
+            low = np.greater(x, self.low)
+
+        if not np.all(low):
+            return False
+
+        if self.high_inclusive:
+            high = np.less_equal(x, self.high)
+        else:
+            high = np.less(x, self.high)
+
+        # Note: np.all returns numpy.bool_
+        return bool(np.all(high))
+
+
+def _inclusive_low_high(interval, dtype=np.float64):
+    """Generate values low and high to be within the interval range.
+
+    This is used in tests only.
+
+    Returns
+    -------
+    low, high : tuple
+        The returned values low and high lie within the interval.
+    """
+    eps = 10 * np.finfo(dtype).eps
+    if interval.low == -np.inf:
+        low = -1e10
+    elif interval.low < 0:
+        low = interval.low * (1 - eps) + eps
+    else:
+        low = interval.low * (1 + eps) + eps
+
+    if interval.high == np.inf:
+        high = 1e10
+    elif interval.high < 0:
+        high = interval.high * (1 + eps) - eps
+    else:
+        high = interval.high * (1 - eps) - eps
+
+    return low, high
+
+
+class BaseLink(ABC):
+    """Abstract base class for differentiable, invertible link functions.
+
+    Convention:
+        - link function g: raw_prediction = g(y_pred)
+        - inverse link h: y_pred = h(raw_prediction)
+
+    For (generalized) linear models, `raw_prediction = X @ coef` is the so
+    called linear predictor, and `y_pred = h(raw_prediction)` is the predicted
+    conditional (on X) expected value of the target `y_true`.
+
+    The methods are not implemented as staticmethods in case a link function needs
+    parameters.
+    """
+
+    is_multiclass = False  # used for testing only
+
+    # Usually, raw_prediction may be any real number and y_pred is an open
+    # interval.
+    # interval_raw_prediction = Interval(-np.inf, np.inf, False, False)
+    interval_y_pred = Interval(-np.inf, np.inf, False, False)
+
+    @abstractmethod
+    def link(self, y_pred, out=None):
+        """Compute the link function g(y_pred).
+
+        The link function maps (predicted) target values to raw predictions,
+        i.e. `g(y_pred) = raw_prediction`.
+
+        Parameters
+        ----------
+        y_pred : array
+            Predicted target values.
+        out : array
+            A location into which the result is stored. If provided, it must
+            have a shape that the inputs broadcast to. If not provided or None,
+            a freshly-allocated array is returned.
+
+        Returns
+        -------
+        out : array
+            Output array, element-wise link function.
+        """
+
+    @abstractmethod
+    def inverse(self, raw_prediction, out=None):
+        """Compute the inverse link function h(raw_prediction).
+
+        The inverse link function maps raw predictions to predicted target
+        values, i.e. `h(raw_prediction) = y_pred`.
+
+        Parameters
+        ----------
+        raw_prediction : array
+            Raw prediction values (in link space).
+        out : array
+            A location into which the result is stored. If provided, it must
+            have a shape that the inputs broadcast to. If not provided or None,
+            a freshly-allocated array is returned.
+
+        Returns
+        -------
+        out : array
+            Output array, element-wise inverse link function.
+        """
+
+
+class IdentityLink(BaseLink):
+    """The identity link function g(x)=x."""
+
+    def link(self, y_pred, out=None):
+        if out is not None:
+            np.copyto(out, y_pred)
+            return out
+        else:
+            return y_pred
+
+    inverse = link
+
+
+class LogLink(BaseLink):
+    """The log link function g(x)=log(x)."""
+
+    interval_y_pred = Interval(0, np.inf, False, False)
+
+    def link(self, y_pred, out=None):
+        return np.log(y_pred, out=out)
+
+    def inverse(self, raw_prediction, out=None):
+        return np.exp(raw_prediction, out=out)
+
+
+class LogitLink(BaseLink):
+    """The logit link function g(x)=logit(x)."""
+
+    interval_y_pred = Interval(0, 1, False, False)
+
+    def link(self, y_pred, out=None):
+        return logit(y_pred, out=out)
+
+    def inverse(self, raw_prediction, out=None):
+        return expit(raw_prediction, out=out)
+
+
+class MultinomialLogit(BaseLink):
+    """The symmetric multinomial logit function.
+
+    Convention:
+        - y_pred.shape = raw_prediction.shape = (n_samples, n_classes)
+
+    Notes:
+        - The inverse link h is the softmax function.
+        - The sum is over the second axis, i.e. axis=1 (n_classes).
+
+    We have to choose additional constraints in order to make
+
+        y_pred[k] = exp(raw_pred[k]) / sum(exp(raw_pred[k]), k=0..n_classes-1)
+
+    for n_classes classes identifiable and invertible.
+    We choose the symmetric side constraint where the geometric mean response
+    is set as reference category, see [2]:
+
+    The symmetric multinomial logit link function for a single data point is
+    then defined as
+
+        raw_prediction[k] = g(y_pred[k]) = log(y_pred[k]/gmean(y_pred))
+        = log(y_pred[k]) - mean(log(y_pred)).
+
+    Note that this is equivalent to the definition in [1] and implies mean
+    centered raw predictions:
+
+        sum(raw_prediction[k], k=0..n_classes-1) = 0.
+
+    For linear models with raw_prediction = X @ coef, this corresponds to
+    sum(coef[k], k=0..n_classes-1) = 0, i.e. the sum over classes for every
+    feature is zero.
+
+    Reference
+    ---------
+    .. [1] Friedman, Jerome; Hastie, Trevor; Tibshirani, Robert. "Additive
+        logistic regression: a statistical view of boosting" Ann. Statist.
+        28 (2000), no. 2, 337--407. doi:10.1214/aos/1016218223.
+        https://projecteuclid.org/euclid.aos/1016218223
+
+    .. [2] Zahid, Faisal Maqbool and Gerhard Tutz. "Ridge estimation for
+        multinomial logit models with symmetric side constraints."
+        Computational Statistics 28 (2013): 1017-1034.
+        http://epub.ub.uni-muenchen.de/11001/1/tr067.pdf
+    """
+
+    is_multiclass = True
+    interval_y_pred = Interval(0, 1, False, False)
+
+    def symmetrize_raw_prediction(self, raw_prediction):
+        return raw_prediction - np.mean(raw_prediction, axis=1)[:, np.newaxis]
+
+    def link(self, y_pred, out=None):
+        # geometric mean as reference category
+        gm = gmean(y_pred, axis=1)
+        return np.log(y_pred / gm[:, np.newaxis], out=out)
+
+    def inverse(self, raw_prediction, out=None):
+        if out is None:
+            return softmax(raw_prediction, copy=True)
+        else:
+            np.copyto(out, raw_prediction)
+            softmax(out, copy=False)
+            return out
+
+
+_LINKS = {
+    "identity": IdentityLink,
+    "log": LogLink,
+    "logit": LogitLink,
+    "multinomial_logit": MultinomialLogit,
+}

dashboard/flask-server/venv/Lib/site-packages/sklearn/_loss/setup.py

@@ -0,0 +1,25 @@
+import numpy
+from numpy.distutils.misc_util import Configuration
+from sklearn._build_utils import gen_from_templates
+
+
+def configuration(parent_package="", top_path=None):
+    config = Configuration("_loss", parent_package, top_path)
+
+    # generate _loss.pyx from template
+    templates = ["sklearn/_loss/_loss.pyx.tp"]
+    gen_from_templates(templates)
+
+    config.add_extension(
+        "_loss",
+        sources=["_loss.pyx"],
+        include_dirs=[numpy.get_include()],
+        # define_macros=[("NPY_NO_DEPRECATED_API", "NPY_1_7_API_VERSION")],
+    )
+    return config
+
+
+if __name__ == "__main__":
+    from numpy.distutils.core import setup
+
+    setup(**configuration().todict())

dashboard/flask-server/venv/Lib/site-packages/sklearn/_loss/tests/test_glm_distribution.py

@@ -0,0 +1,123 @@
+# Authors: Christian Lorentzen <lorentzen.ch@gmail.com>
+#
+# License: BSD 3 clause
+#
+# TODO(1.3): remove file
+import numpy as np
+from numpy.testing import (
+    assert_allclose,
+    assert_array_equal,
+)
+from scipy.optimize import check_grad
+import pytest
+
+from sklearn._loss.glm_distribution import (
+    TweedieDistribution,
+    NormalDistribution,
+    PoissonDistribution,
+    GammaDistribution,
+    InverseGaussianDistribution,
+    DistributionBoundary,
+)
+
+
+@pytest.mark.parametrize(
+    "family, expected",
+    [
+        (NormalDistribution(), [True, True, True]),
+        (PoissonDistribution(), [False, True, True]),
+        (TweedieDistribution(power=1.5), [False, True, True]),
+        (GammaDistribution(), [False, False, True]),
+        (InverseGaussianDistribution(), [False, False, True]),
+        (TweedieDistribution(power=4.5), [False, False, True]),
+    ],
+)
+def test_family_bounds(family, expected):
+    """Test the valid range of distributions at -1, 0, 1."""
+    result = family.in_y_range([-1, 0, 1])
+    assert_array_equal(result, expected)
+
+
+def test_invalid_distribution_bound():
+    dist = TweedieDistribution()
+    dist._lower_bound = 0
+    with pytest.raises(TypeError, match="must be of type DistributionBoundary"):
+        dist.in_y_range([-1, 0, 1])
+
+
+def test_tweedie_distribution_power():
+    msg = "distribution is only defined for power<=0 and power>=1"
+    with pytest.raises(ValueError, match=msg):
+        TweedieDistribution(power=0.5)
+
+    with pytest.raises(TypeError, match="must be a real number"):
+        TweedieDistribution(power=1j)
+
+    with pytest.raises(TypeError, match="must be a real number"):
+        dist = TweedieDistribution()
+        dist.power = 1j
+
+    dist = TweedieDistribution()
+    assert isinstance(dist._lower_bound, DistributionBoundary)
+
+    assert dist._lower_bound.inclusive is False
+    dist.power = 1
+    assert dist._lower_bound.value == 0.0
+    assert dist._lower_bound.inclusive is True
+
+
+@pytest.mark.parametrize(
+    "family, chk_values",
+    [
+        (NormalDistribution(), [-1.5, -0.1, 0.1, 2.5]),
+        (PoissonDistribution(), [0.1, 1.5]),
+        (GammaDistribution(), [0.1, 1.5]),
+        (InverseGaussianDistribution(), [0.1, 1.5]),
+        (TweedieDistribution(power=-2.5), [0.1, 1.5]),
+        (TweedieDistribution(power=-1), [0.1, 1.5]),
+        (TweedieDistribution(power=1.5), [0.1, 1.5]),
+        (TweedieDistribution(power=2.5), [0.1, 1.5]),
+        (TweedieDistribution(power=-4), [0.1, 1.5]),
+    ],
+)
+def test_deviance_zero(family, chk_values):
+    """Test deviance(y,y) = 0 for different families."""
+    for x in chk_values:
+        assert_allclose(family.deviance(x, x), 0, atol=1e-9)
+
+
+@pytest.mark.parametrize(
+    "family",
+    [
+        NormalDistribution(),
+        PoissonDistribution(),
+        GammaDistribution(),
+        InverseGaussianDistribution(),
+        TweedieDistribution(power=-2.5),
+        TweedieDistribution(power=-1),
+        TweedieDistribution(power=1.5),
+        TweedieDistribution(power=2.5),
+        TweedieDistribution(power=-4),
+    ],
+    ids=lambda x: x.__class__.__name__,
+)
+def test_deviance_derivative(family):
+    """Test deviance derivative for different families."""
+    rng = np.random.RandomState(0)
+    y_true = rng.rand(10)
+    # make data positive
+    y_true += np.abs(y_true.min()) + 1e-2
+
+    y_pred = y_true + np.fmax(rng.rand(10), 0.0)
+
+    dev = family.deviance(y_true, y_pred)
+    assert isinstance(dev, float)
+    dev_derivative = family.deviance_derivative(y_true, y_pred)
+    assert dev_derivative.shape == y_pred.shape
+
+    err = check_grad(
+        lambda y_pred: family.deviance(y_true, y_pred),
+        lambda y_pred: family.deviance_derivative(y_true, y_pred),
+        y_pred,
+    ) / np.linalg.norm(dev_derivative)
+    assert abs(err) < 1e-6

dashboard/flask-server/venv/Lib/site-packages/sklearn/_loss/tests/test_link.py

@@ -0,0 +1,108 @@
+import numpy as np
+from numpy.testing import assert_allclose, assert_array_equal
+import pytest
+
+from sklearn._loss.link import (
+    _LINKS,
+    _inclusive_low_high,
+    MultinomialLogit,
+    Interval,
+)
+
+
+LINK_FUNCTIONS = list(_LINKS.values())
+
+
+def test_interval_raises():
+    """Test that interval with low > high raises ValueError."""
+    with pytest.raises(
+        ValueError, match="One must have low <= high; got low=1, high=0."
+    ):
+        Interval(1, 0, False, False)
+
+
+@pytest.mark.parametrize(
+    "interval",
+    [
+        Interval(0, 1, False, False),
+        Interval(0, 1, False, True),
+        Interval(0, 1, True, False),
+        Interval(0, 1, True, True),
+        Interval(-np.inf, np.inf, False, False),
+        Interval(-np.inf, np.inf, False, True),
+        Interval(-np.inf, np.inf, True, False),
+        Interval(-np.inf, np.inf, True, True),
+        Interval(-10, -1, False, False),
+        Interval(-10, -1, False, True),
+        Interval(-10, -1, True, False),
+        Interval(-10, -1, True, True),
+    ],
+)
+def test_is_in_range(interval):
+    # make sure low and high are always within the interval, used for linspace
+    low, high = _inclusive_low_high(interval)
+
+    x = np.linspace(low, high, num=10)
+    assert interval.includes(x)
+
+    # x contains lower bound
+    assert interval.includes(np.r_[x, interval.low]) == interval.low_inclusive
+
+    # x contains upper bound
+    assert interval.includes(np.r_[x, interval.high]) == interval.high_inclusive
+
+    # x contains upper and lower bound
+    assert interval.includes(np.r_[x, interval.low, interval.high]) == (
+        interval.low_inclusive and interval.high_inclusive
+    )
+
+
+@pytest.mark.parametrize("link", LINK_FUNCTIONS)
+def test_link_inverse_identity(link):
+    # Test that link of inverse gives identity.
+    rng = np.random.RandomState(42)
+    link = link()
+    n_samples, n_classes = 100, None
+    if link.is_multiclass:
+        n_classes = 10
+        raw_prediction = rng.normal(loc=0, scale=10, size=(n_samples, n_classes))
+        if isinstance(link, MultinomialLogit):
+            raw_prediction = link.symmetrize_raw_prediction(raw_prediction)
+    else:
+        # So far, the valid interval of raw_prediction is (-inf, inf) and
+        # we do not need to distinguish.
+        raw_prediction = rng.normal(loc=0, scale=10, size=(n_samples))
+
+    assert_allclose(link.link(link.inverse(raw_prediction)), raw_prediction)
+    y_pred = link.inverse(raw_prediction)
+    assert_allclose(link.inverse(link.link(y_pred)), y_pred)
+
+
+@pytest.mark.parametrize("link", LINK_FUNCTIONS)
+def test_link_out_argument(link):
+    # Test that out argument gets assigned the result.
+    rng = np.random.RandomState(42)
+    link = link()
+    n_samples, n_classes = 100, None
+    if link.is_multiclass:
+        n_classes = 10
+        raw_prediction = rng.normal(loc=0, scale=10, size=(n_samples, n_classes))
+        if isinstance(link, MultinomialLogit):
+            raw_prediction = link.symmetrize_raw_prediction(raw_prediction)
+    else:
+        # So far, the valid interval of raw_prediction is (-inf, inf) and
+        # we do not need to distinguish.
+        raw_prediction = rng.normal(loc=0, scale=10, size=(n_samples))
+
+    y_pred = link.inverse(raw_prediction, out=None)
+    out = np.empty_like(raw_prediction)
+    y_pred_2 = link.inverse(raw_prediction, out=out)
+    assert_allclose(y_pred, out)
+    assert_array_equal(out, y_pred_2)
+    assert np.shares_memory(out, y_pred_2)
+
+    out = np.empty_like(y_pred)
+    raw_prediction_2 = link.link(y_pred, out=out)
+    assert_allclose(raw_prediction, out)
+    assert_array_equal(out, raw_prediction_2)
+    assert np.shares_memory(out, raw_prediction_2)