DeepXDE v0.13.2

• Backend PyTorch supports PDE forward problems

API changes

• Refactor uncertainty via MC dropout as the callback DropoutUncertainty; remove "uncertainty" argument from Model.train().

DeepXDE v0.13.1

DeepXDE supports PyTorch backend for function approximation.

Areas of improvement

• Backend TensorFlow supports auxiliary variables

New APIs

• Add dde.config.set_default_float() and dde.config.default_float()

DeepXDE v0.13.0

DeepXDE now supports two backends: TensorFlow 1.x (tensorflow.compat.v1 in TensorFlow 2.x) and TensorFlow 2.x. For how to select one, see Working with different backends.

Areas of improvement

• Many modules are refactored to better support multiple backends.
• Support skopt>=0.9
• Documentation improvements

API changes

• Rename dde.data.Func to dde.data.Function

New APIs

• Add Hypercube.random_boundary_points()

DeepXDE v0.12.0

This release is mainly about DeepONet.

API changes

• Rename OpNN to DeepONet
• Rename OpDataSet to Triple

New APIs

• Add dde.__version__
• Add data.TripleCartesianProd, maps.DeepONetCartesianProd, and maps.FourierDeepONetCartesianProd
• Add new metric: mean_l2_relative_error

Areas of improvement

• Bug fix: change 'sobol' to 'Sobol'

DeepXDE v0.11.2

Areas of improvement

• Add Multi-scale Fourier Feature Neural Networks: MsFFN and STMsFFN
• PDE supports more sampling methods: LHS, Halton, Hammersley
• DeepONet supports input_transform and output_transform
• PointSet supports default value
• Hypercube.boundary_normal() returns averaged normal for vertices
• Speedup Polygon.random_points()

DeepXDE v0.11.1

Areas of improvement

• FNN supports arguments use_bias and kernel_constraint, and layer normalization
• Change L-BFGS option gtol from 1e-5 to 1e-8
• Improve saveplot

New APIs

• Add a new Data Constraint
• Add metric mean_squared_error

DeepXDE v0.11.0

We stop the support of Python 2.7 from this release.

Areas of improvement

• Vectorize geometry related methods
• PDE supports resample PDE residual points via callback PDEResidualResampler
• Implement Polygon.boundary_normal()

New APIs

• Add multifidelity DeepONet

API changes

• Geometry-related methods (inside, on_boundary, on_initial, boundary_normal, periodic_point) now use the 2d input of shape N by d, not a 1d vector.

DeepXDE v0.10.0

Areas of improvement

• PDE supports excluded points
• Add PointSetBC for training using exact solution.
• Add Hypersphere.boundary_normal()
• Add Triangle.inside()

API changes

• PDE doesn't support numpy.array X as the third parameter. To define pde with extra coefficients, use PDE(..., auxiliary_var_function=...), see an example at Lorenz_inverse_forced_Colab.ipynb.

DeepXDE v0.9.1

Areas of improvement

• Model.predict() supports PDE with two or three arguments
• PDE supports different loss functions for different errors
• Hypercube.uniform_points() uses the same dx in each dimension

New APIs

• Add metric: nanl2_relative_error, zero
• Add callback: Timer

DeepXDE v0.9.0

New module

• Add new module dde.grad to compute gradients via dde.grad.jacobian() and dde.grad.hessian()

Areas of improvement

• Bug fix: GeometryXTime.periodic_point()