TensorFlow 2.0 Beta is available Learn more

Module: tfp.bijectors

Bijective transformations.

Defined in python/bijectors/__init__.py.


class AbsoluteValue: Computes Y = g(X) = Abs(X), element-wise.

class Affine: Compute Y = g(X; shift, scale) = scale @ X + shift.

class AffineLinearOperator: Compute Y = g(X; shift, scale) = scale @ X + shift.

class AffineScalar: Compute Y = g(X; shift, scale) = scale * X + shift.

class AutoregressiveLayer: Masked Autoencoder for Distribution Estimation [Germain et al. (2015)].

class AutoregressiveNetwork: Masked Autoencoder for Distribution Estimation [Germain et al. (2015)][1].

class BatchNormalization: Compute Y = g(X) s.t. X = g^-1(Y) = (Y - mean(Y)) / std(Y).

class Bijector: Interface for transformations of a Distribution sample.

class Blockwise: Bijector which applies a list of bijectors to blocks of a Tensor.

class Chain: Bijector which applies a sequence of bijectors.

class CholeskyOuterProduct: Compute g(X) = X @ X.T; X is lower-triangular, positive-diagonal matrix.

class CholeskyToInvCholesky: Maps the Cholesky factor of M to the Cholesky factor of M^{-1}.

class ConditionalBijector: Conditional Bijector is a Bijector that allows intrinsic conditioning.

class Cumsum: Computes the cumulative sum of a tensor along a specified axis.

class DiscreteCosineTransform: Compute Y = g(X) = DCT(X), where DCT type is indicated by the type arg.

class Exp: Compute Y = g(X) = exp(X).

class Expm1: Compute Y = g(X) = exp(X) - 1.

class FillTriangular: Transforms vectors to triangular.

class Gumbel: Compute Y = g(X) = exp(-exp(-(X - loc) / scale)).

class Identity: Compute Y = g(X) = X.

class Inline: Bijector constructed from custom callables.

class Invert: Bijector which inverts another Bijector.

class IteratedSigmoidCentered: Bijector which applies a Stick Breaking procedure.

class Kumaraswamy: Compute Y = g(X) = (1 - (1 - X)**(1 / b))**(1 / a), X in [0, 1].

class MaskedAutoregressiveFlow: Affine MaskedAutoregressiveFlow bijector.

class MatrixInverseTriL: Computes g(L) = inv(L), where L is a lower-triangular matrix.

class MatvecLU: Matrix-vector multiply using LU decomposition.

class NormalCDF: Compute Y = g(X) = NormalCDF(x).

class Ordered: Bijector which maps a tensor x_k that has increasing elements in the last

class Permute: Permutes the rightmost dimension of a Tensor.

class PowerTransform: Compute Y = g(X) = (1 + X * c)**(1 / c), X >= -1 / c.

class RealNVP: RealNVP "affine coupling layer" for vector-valued events.

class Reciprocal: A Bijector that computes the reciprocal b(x) = 1. / x entrywise.

class Reshape: Reshapes the event_shape of a Tensor.

class ScaleTriL: Transforms unconstrained vectors to TriL matrices with positive diagonal.

class Sigmoid: Bijector which computes Y = g(X) = 1 / (1 + exp(-X)).

class SinhArcsinh: Compute Y = g(X) = Sinh( (Arcsinh(X) + skewness) * tailweight ).

class SoftmaxCentered: Bijector which computes Y = g(X) = exp([X 0]) / sum(exp([X 0])).

class Softplus: Bijector which computes Y = g(X) = Log[1 + exp(X)].

class Softsign: Bijector which computes Y = g(X) = X / (1 + |X|).

class Square: Compute g(X) = X^2; X is a positive real number.

class Tanh: Bijector that computes Y = tanh(X), therefore Y in (-1, 1).

class TransformDiagonal: Applies a Bijector to the diagonal of a matrix.

class Transpose: Compute Y = g(X) = transpose_rightmost_dims(X, rightmost_perm).

class Weibull: Compute Y = g(X) = 1 - exp((-X / scale) ** concentration), X >= 0.


masked_autoregressive_default_template(...): Build the Masked Autoregressive Density Estimator (Germain et al., 2015).

masked_dense(...): A autoregressively masked dense layer. Analogous to tf.layers.dense.

real_nvp_default_template(...): Build a scale-and-shift function using a multi-layer neural network.