# Copyright 2016 James Hensman, alexggmatthews, Mark van der Wilk
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from typing import Optional

import numpy as np
import tensorflow as tf

from gpflow.kernels import Kernel
from .model import MeanAndVariance, GPModel, Data, GPPosterior
from ..likelihoods import Likelihood, Gaussian
from ..config import default_float, default_jitter
from ..covariances.dispatch import Kuf, Kuu
from ..inducing_variables import InducingPoints
from ..mean_functions import Zero, MeanFunction
from .util import inducingpoint_wrapper


class SGPR(GPModel):
    """
    Sparse Variational GP regression. The key reference is

    ::

      @inproceedings{titsias2009variational,
        title={Variational learning of inducing variables in
               sparse Gaussian processes},
        author={Titsias, Michalis K},
        booktitle={International Conference on
                   Artificial Intelligence and Statistics},
        pages={567--574},
        year={2009}
      }
    """

    def __init__(self,
                 kernel: Kernel,
                 inducing_variable: InducingPoints,
                 mean_function: Optional[MeanFunction] = None,
                 likelihood: Optional[Likelihood] = None
                 ) -> None:
        """
        Z is a matrix of pseudo inputs, size [M, D]
        kernel, mean_function are appropriate GPflow objects

        This method only works with a Gaussian likelihood.
        """
        if likelihood is None:
            likelihood = Gaussian(variance=1.0)
        elif not isinstance(likelihood, Gaussian):
            raise ValueError("Likelihood must be Gaussian to work with SGPR")
        super().__init__(kernel, likelihood, mean_function)
        self.inducing_variable = inducingpoint_wrapper(inducing_variable)

    def _common_terms(self, data: Data):
        x_data, y_data = data
        num_inducing = len(self.inducing_variable)

        err = y_data - self.mean_function(x_data)
        Kdiag = self.kernel.K_diag(x_data)
        kuf = Kuf(self.inducing_variable, self.kernel, x_data)
        kuu = Kuu(self.inducing_variable, self.kernel, jitter=default_jitter())
        Luu = tf.linalg.cholesky(kuu)
        sigma = tf.sqrt(self.likelihood.variance)

        # Compute intermediate matrices
        A = tf.linalg.triangular_solve(Luu, kuf, lower=True) / sigma
        AAT = tf.linalg.matmul(A, A, transpose_b=True)
        B = AAT + tf.eye(num_inducing, dtype=default_float())
        LB = tf.linalg.cholesky(B)
        Aerr = tf.linalg.matmul(A, err)
        c = tf.linalg.triangular_solve(LB, Aerr, lower=True) / sigma

        trace_term = -0.5 * tf.reduce_sum(Kdiag) / self.likelihood.variance \
                     +0.5 * tf.reduce_sum(tf.linalg.diag_part(AAT))

        return err, B, c, Luu, LB, trace_term

    def elbo(self, data: Data):
        """
        data is a tuple of X, Y with
          X is a data matrix, size [N, D]
          Y is a data matrix, size [N, R]

        Construct a tensorflow function to compute the bound on the marginal
        likelihood. For a derivation of the terms in here, see the associated
        SGPR notebook.
        """
        err, B, c, Luu, LB, trace_term = self._common_terms(data)

        output_dim = tf.cast(tf.shape(data[1])[1], default_float())
        num_data = tf.cast(tf.shape(data[0])[0], default_float())

        bound = -0.5 * num_data * output_dim * np.log(2 * np.pi)
        bound -= output_dim * tf.reduce_sum(tf.math.log(tf.linalg.diag_part(LB)))
        bound -= 0.5 * num_data * output_dim * tf.math.log(self.likelihood.variance)
        bound -= 0.5 * tf.reduce_sum(tf.square(err)) / self.likelihood.variance
        bound += 0.5 * tf.reduce_sum(tf.square(c))
        bound += output_dim * trace_term

        return bound

    def objective(self, data: Data):
        return -self.elbo(data)

    def optimize(self):
        pass
        # opt = trainig.Scipy()
        # with tf.GradientTape(...):
        #     pass
            # TODO

    def get_posterior(self, data: Data):
        # 1: compute q_mu and q_sqrt for the given data
        _, B, c, Luu, LB, _ = self._common_terms(data)

        # TODO: this might be able to be more effective if we could pass in
        # chol(K) or equavalent to conditionals (see also:GPR)
        q_mu = tf.linalg.triangular_solve(tf.transpose(LB), c, lower=False)
        LBi = tf.linalg.inv(LB)
        Bi = tf.matmul(LBi, LBi, transpose_a=True)
        num_inducing = len(self.inducing_variable)
        # variance_sqrt = tf.linalg.cholesky(tf.linalg.inv(B) + tf.eye(num_inducing, dtype=default_float())*1e-6)
        variance_sqrt = tf.linalg.cholesky(Bi + tf.eye(num_inducing, dtype=default_float())*1e-12)

        # TODO: make copies?
        # 2: create posterior object
        return GPPosterior(mean_function=self.mean_function,
                           kernel=self.kernel,
                           likelihood=self.likelihood,
                           inducing_variable=self.inducing_variable,
                           whiten=True,
                           mean=q_mu.numpy(),
                           variance_sqrt=variance_sqrt.numpy())


class GPRFITC(GPModel):
    def __init__(self,
                 kernel: Kernel,
                 inducing_variable: InducingPoints,
                 mean_function: Optional[MeanFunction] = None,
                 likelihood: Optional[Likelihood] = None
                 ):
        """
        This implements GP regression with the FITC approximation.
        The key reference is

        @inproceedings{Snelson06sparsegaussian,
        author = {Edward Snelson and Zoubin Ghahramani},
        title = {Sparse Gaussian Processes using Pseudo-inputs},
        booktitle = {Advances In Neural Information Processing Systems },
        year = {2006},
        pages = {1257--1264},
        publisher = {MIT press}
        }

        Implementation loosely based on code from GPML matlab library although
        obviously gradients are automatic in GPflow.

        X is a data matrix, size [N, D]
        Y is a data matrix, size [N, R]
        Z is a matrix of pseudo inputs, size [M, D]
        kernel, mean_function are appropriate GPflow objects

        This method only works with a Gaussian likelihood.

        """

        if likelihood is None:
            likelihood = Gaussian(variance=1.0)
        elif not isinstance(likelihood, Gaussian):
            raise ValueError("Likelihood must be Gaussian to work with FITC")
        super().__init__(kernel, likelihood, mean_function)
        self.inducing_variable = inducingpoint_wrapper(inducing_variable)

    def _common_terms(self, data: Data):
        x_data, y_data = data
        num_inducing = len(self.inducing_variable)
        err = y_data - self.mean_function(x_data)  # size [N, R]
        Kdiag = self.kernel.K_diag(x_data)
        kuf = Kuf(self.inducing_variable, self.kernel, x_data)
        kuu = Kuu(self.inducing_variable, self.kernel, jitter=default_jitter())

        Luu = tf.linalg.cholesky(kuu)  # => Luu Luu^T = kuu
        V = tf.linalg.triangular_solve(
            Luu, kuf)  # => V^T V = Qff = kuf^T kuu^-1 kuf

        diagQff = tf.reduce_sum(tf.square(V), 0)
        nu = Kdiag - diagQff + self.likelihood.variance

        B = tf.eye(num_inducing, dtype=default_float()) + tf.linalg.matmul(
            V / nu, V, transpose_b=True)
        L = tf.linalg.cholesky(B)
        beta = err / tf.expand_dims(nu, 1)  # size [N, R]
        alpha = tf.linalg.matmul(V, beta)  # size [N, R]

        gamma = tf.linalg.triangular_solve(L, alpha, lower=True)  # size [N, R]

        return err, nu, Luu, L, alpha, beta, gamma

    def log_likelihood(self, data: Data) -> tf.Tensor:
        """
        Construct a tensorflow function to compute the bound on the marginal
        likelihood.
        """

        # FITC approximation to the log marginal likelihood is
        # log ( normal( y | mean, K_fitc ) )
        # where K_fitc = Qff + diag( \nu )
        # where Qff = Kfu kuu^{-1} kuf
        # with \nu_i = Kff_{i,i} - Qff_{i,i} + \sigma^2

        # We need to compute the Mahalanobis term -0.5* err^T K_fitc^{-1} err
        # (summed over functions).

        # We need to deal with the matrix inverse term.
        # K_fitc^{-1} = ( Qff + \diag( \nu ) )^{-1}
        #            = ( V^T V + \diag( \nu ) )^{-1}
        # Applying the Woodbury identity we obtain
        #            = \diag( \nu^{-1} ) - \diag( \nu^{-1} ) V^T ( I + V \diag( \nu^{-1} ) V^T )^{-1) V \diag(\nu^{-1} )
        # Let \beta =  \diag( \nu^{-1} ) err
        # and let \alpha = V \beta
        # then Mahalanobis term = -0.5* ( \beta^T err - \alpha^T Solve( I + V \diag( \nu^{-1} ) V^T, alpha ) )

        err, nu, Luu, L, alpha, beta, gamma = self._common_terms(data)

        mahalanobisTerm = -0.5 * tf.reduce_sum(tf.square(err) / tf.expand_dims(nu, 1)) \
                          + 0.5 * tf.reduce_sum(tf.square(gamma))

        # We need to compute the log normalizing term -N/2 \log 2 pi - 0.5 \log \det( K_fitc )

        # We need to deal with the log determinant term.
        # \log \det( K_fitc ) = \log \det( Qff + \diag( \nu ) )
        #                    = \log \det( V^T V + \diag( \nu ) )
        # Applying the determinant lemma we obtain
        #                    = \log [ \det \diag( \nu ) \det( I + V \diag( \nu^{-1} ) V^T ) ]
        #                    = \log [ \det \diag( \nu ) ] + \log [ \det( I + V \diag( \nu^{-1} ) V^T ) ]

        num_data = data[0].shape[0]
        num_latent = data[1].shape[1]
        constantTerm = -0.5 * num_data * tf.math.log(
            tf.constant(2. * np.pi, default_float()))
        logDeterminantTerm = -0.5 * tf.reduce_sum(
            tf.math.log(nu)) - tf.reduce_sum(
            tf.math.log(tf.linalg.diag_part(L)))
        logNormalizingTerm = constantTerm + logDeterminantTerm

        return mahalanobisTerm + logNormalizingTerm * num_latent

    def objective(self, data: Data) -> tf.Tensor:
        return -self.log_likelihood(data)

    def get_posterior(self, data: Data):
        _, _, Luu, L, _, _, gamma = self._common_terms(data)
        q_mu = tf.linalg.solve(L, gamma)
        variance_sqrt = L

        return GPPosterior(mean_function=self.mean_function,
                           kernel=self.kernel,
                           likelihood=self.likelihood,
                           inducing_variable=self.inducing_variable,
                           whiten=True,
                           mean=q_mu.numpy(),
                           variance_sqrt=variance_sqrt.numpy())




    def predict_f(self, X: tf.Tensor, full_cov=False,
                  full_output_cov=False) -> MeanAndVariance:
        """
        Compute the mean and variance of the latent function at some new points
        Xnew.
        """
        _, _, Luu, L, _, _, gamma = self.common_terms()
        Kus = Kuf(self.inducing_variable, self.kernel, X)  # size  [M, X]new

        w = tf.linalg.triangular_solve(Luu, Kus, lower=True)  # size [M, X]new

        tmp = tf.linalg.triangular_solve(tf.transpose(L), gamma, lower=False)
        mean = tf.linalg.matmul(w, tmp,
                                transpose_a=True) + self.mean_function(X)
        intermediateA = tf.linalg.triangular_solve(L, w, lower=True)

        if full_cov:
            var = self.kernel(X) - tf.linalg.matmul(w, w, transpose_a=True) \
                  + tf.linalg.matmul(intermediateA, intermediateA, transpose_a=True)
            var = tf.tile(var[None, ...], [num_latent, 1, 1])  # [P, N, N]
        else:
            var = self.kernel(X, full=False) - tf.reduce_sum(tf.square(w), 0) \
                  + tf.reduce_sum(tf.square(intermediateA), 0)  # size Xnew,
            var = tf.tile(var[:, None], [1, num_latent])

        return mean, var



def upper_bound_for_sgpr_models(model, data: Data):
    """
    Upper bound for the GP regression marginal likelihood.
    It is implemented here as a Mixin class which works with SGPR and GPRFITC.
    Note that the same inducing points are used for calculating the upper bound,
    as are used for computing the likelihood approximation. This may not lead to
    the best upper bound. The upper bound can be tightened by optimising Z, just
    as just like the lower bound. This is especially important in FITC, as FITC
    is known to produce poor inducing point locations. An optimisable upper bound
    can be found in https://github.com/markvdw/gp_upper.

    The key reference is

    ::

    @misc{titsias_2014,
        title={Variational Inference for Gaussian and Determinantal Point Processes},
        url={http://www2.aueb.gr/users/mtitsias/papers/titsiasNipsVar14.pdf},
        publisher={Workshop on Advances in Variational Inference (NIPS 2014)},
        author={Titsias, Michalis K.},
        year={2014},
        month={Dec}
    }
    """
    if not isinstance(model, (SGPR, GPRFITC)):
        raise ValueError("The upper bound can only be computed "
                         "for models of the type SGPR or GPRFITC")

    x_data, y_data = data
    num_data = tf.cast(tf.shape(y_data)[0], default_float())

    Kdiag = model.kernel(x_data)
    kuu = Kuu(model.inducing_variable, model.kernel, jitter=default_jitter())
    kuf = Kuf(model.inducing_variable, model.kernel, x_data)

    L = tf.linalg.cholesky(kuu)
    LB = tf.linalg.cholesky(kuu + model.likelihood.variance ** -1.0 *
                            tf.linalg.matmul(kuf, kuf, transpose_b=True))

    Linvkuf = tf.linalg.triangular_solve(L, kuf, lower=True)
    # Using the Trace bound, from Titsias' presentation
    # c = tf.reduce_sum(Kdiag) - tf.reduce_sum(Linvkuf ** 2.0)
    Kff = model.kernel(x_data)
    Qff = tf.linalg.matmul(kuf, Linvkuf, transpose_a=True)

    # Alternative bound on max eigenval:
    c = tf.reduce_max(tf.reduce_sum(tf.abs(Kff - Qff), 0))
    corrected_noise = model.likelihood.variance + c

    const = -0.5 * num_data * tf.math.log(
        2 * np.pi * model.likelihood.variance)
    logdet = tf.reduce_sum(tf.math.log(
        tf.linalg.diag_part(L))) - tf.reduce_sum(
        tf.math.log(tf.linalg.diag_part(LB)))

    LC = tf.linalg.cholesky(kuu + corrected_noise ** -1.0 *
                            tf.linalg.matmul(kuf, kuf, transpose_b=True))
    v = tf.linalg.triangular_solve(LC,
                                    corrected_noise ** -1.0 *
                                    tf.linalg.matmul(kuf, y_data),
                                    lower=True)
    quad = -0.5 * corrected_noise ** -1.0 * tf.reduce_sum(
        y_data ** 2.0) + 0.5 * tf.reduce_sum(v ** 2.0)

    return const + logdet + quad