v5.18.1

NeuralPDE v5.18.1

Diff since v5.18.0

Merged pull requests:

• CompatHelper: bump compat for AdvancedHMC to 0.7, (keep existing compat) (#926) (@github-actions[bot])
• CompatHelper: bump compat for AdvancedHMC to 0.7 for package docs, (keep existing compat) (#927) (@github-actions[bot])
• docs: fix pino tutorial (#930) (@sathvikbhagavan)

Closed issues:

• NeuralPDE gives a MLDataDevice.UnknownDevice error even though the code is entirely executing on the CPU (#922)

v5.18.0

NeuralPDE v5.18.0

Diff since v5.17.0

Merged pull requests:

• Physics informed neural operator ode (#806) (@KirillZubov)
• improved BPINN solvers (#905) (@AstitvaAggarwal)
• CompatHelper: bump compat for Functors to 0.5, (keep existing compat) (#907) (@github-actions[bot])
• build(deps): bump crate-ci/typos from 1.26.0 to 1.27.0 (#908) (@dependabot[bot])
• CompatHelper: bump compat for Optimisers to 0.4, (keep existing compat) (#909) (@github-actions[bot])
• Real valued datagen from GridTraining when boundary points are trivial (#910) (@ChrisRackauckas)
• build(deps): bump codecov/codecov-action from 4 to 5 (#912) (@dependabot[bot])
• CompatHelper: bump compat for Flux to 0.15 for package docs, (keep existing compat) (#915) (@github-actions[bot])
• build(deps): bump crate-ci/typos from 1.27.0 to 1.28.2 (#916) (@dependabot[bot])
• CompatHelper: bump compat for Flux to 0.16 for package docs, (keep existing compat) (#917) (@github-actions[bot])
• CompatHelper: bump compat for Zygote to 0.7, (keep existing compat) (#918) (@github-actions[bot])
• build(deps): bump crate-ci/typos from 1.28.2 to 1.29.4 (#919) (@dependabot[bot])
• Fix CI (#921) (@sathvikbhagavan)
• build(deps): bump crate-ci/typos from 1.29.4 to 1.30.0 (#925) (@dependabot[bot])

Closed issues:

• GSoCs Assemble! (#670)
• Incorrect adapt usage leads to non-concrete elements (#883)
• GridTraining generates complex training sets (#906)

v5.17.0

NeuralPDE v5.17.0

Diff since v5.16.0

Merged pull requests:

• build(deps): bump crate-ci/typos from 1.19.0 to 1.20.4 (#847) (@dependabot[bot])
• build(deps): bump crate-ci/typos from 1.20.4 to 1.20.8 (#848) (@dependabot[bot])
• Docstring typo fixes (#849) (@lanceXwq)
• build(deps): bump crate-ci/typos from 1.20.8 to 1.20.9 (#850) (@dependabot[bot])
• build(deps): bump crate-ci/typos from 1.20.9 to 1.20.10 (#851) (@dependabot[bot])
• build(deps): bump crate-ci/typos from 1.20.10 to 1.21.0 (#852) (@dependabot[bot])
• Better BPINN ode Solver (#853) (@AstitvaAggarwal)
• CompatHelper: bump compat for OptimizationOptimJL to 0.3 for package docs, (keep existing compat) (#856) (@github-actions[bot])
• CompatHelper: bump compat for SymbolicUtils to 2, (keep existing compat) (#858) (@github-actions[bot])
• build(deps): bump crate-ci/typos from 1.21.0 to 1.22.0 (#859) (@dependabot[bot])
• build(deps): bump crate-ci/typos from 1.22.0 to 1.22.3 (#863) (@dependabot[bot])
• build(deps): bump crate-ci/typos from 1.22.3 to 1.22.7 (#865) (@dependabot[bot])
• build(deps): bump crate-ci/typos from 1.22.7 to 1.22.9 (#868) (@dependabot[bot])
• chore: fix typos (#869) (@sathvikbhagavan)
• build: ignore patch versions of typos gha in dependabot (#870) (@sathvikbhagavan)
• fix: checking complex type in the parameters of nn (#872) (@sathvikbhagavan)
• build(deps): bump crate-ci/typos from 1.22.9 to 1.23.1 (#873) (@dependabot[bot])
• CompatHelper: bump compat for SymbolicUtils to 3, (keep existing compat) (#878) (@github-actions[bot])
• ci: standardize workflows using SciML's reusable workflows (#879) (@thazhemadam)
• CompatHelper: bump compat for Symbolics to 6, (keep existing compat) (#880) (@github-actions[bot])
• refactor: trigger build with latest Lux (#882) (@avik-pal)
• build(deps): bump crate-ci/typos from 1.23.1 to 1.24.1 (#884) (@dependabot[bot])
• CompatHelper: bump compat for Optimization to 4, (keep existing compat) (#888) (@github-actions[bot])
• CompatHelper: bump compat for Optimization to 4 for package docs, (keep existing compat) (#889) (@github-actions[bot])
• CompatHelper: bump compat for OptimizationOptimisers to 0.3, (keep existing compat) (#890) (@github-actions[bot])
• CompatHelper: bump compat for OptimizationOptimisers to 0.3 for package docs, (keep existing compat) (#891) (@github-actions[bot])
• CompatHelper: bump compat for OptimizationOptimJL to 0.4 for package docs, (keep existing compat) (#892) (@github-actions[bot])
• build(deps): bump crate-ci/typos from 1.24.1 to 1.25.0 (#896) (@dependabot[bot])
• build(deps): bump crate-ci/typos from 1.25.0 to 1.26.0 (#899) (@dependabot[bot])
• refactor: remove NeuralPDELogging in-favor of extension (#901) (@avik-pal)
• ci: try taming down CI timings (#903) (@avik-pal)
• ci: test with 1, lts and pre versions of julia (#904) (@thazhemadam)

Closed issues:

• Iteration should be a Ref{Int}, not a size 1 Vector{Int} (#552)
• Tutorial: imposing constraint on PINN solution of Fokker Planck fails (#854)
• Error when running Front Page Example Code (#881)

v5.16.0

NeuralPDE v5.16.0

Diff since v5.15.0

Merged pull requests:

• chore: format the whole repo (#841) (@sathvikbhagavan)
• build(deps): bump julia-actions/setup-julia from 1 to 2 (#843) (@dependabot[bot])
• ci: add NNODE, NeuralAdapter, IntegroDiff in GHA from buildkite to have them in Downgrade CI (#845) (@sathvikbhagavan)
• Tag the original solution to sol.original and simplify dependencies (#846) (@ChrisRackauckas)

Closed issues:

• Neural Adaptor tests set as @test_broken (#513)
• Multithreaded test suite (#706)
• Add support for equations with complex numbers (#821)

v5.15.0

NeuralPDE v5.15.0

Diff since v5.14.0

Merged pull requests:

• build: bump mtk to v9 (#828) (@sathvikbhagavan)
• ci: add Downgrade CI (#831) (@sathvikbhagavan)
• docs: fix rendering (#832) (@sathvikbhagavan)
• refactor: correctly lower quadrature training strategy in NNODE (#836) (@sathvikbhagavan)
• feat: allow complex for NNODE (#839) (@sathvikbhagavan)

Closed issues:

• Allow batching with QuadratureStrategy in NNODE (#555)

v5.14.0

NeuralPDE v5.14.0

Diff since v5.13.0

Merged pull requests:

• docs: update and cleanup tutorials and examples (#815) (@sathvikbhagavan)

Closed issues:

• Neural Adapter Documentation Example Not Working (#824)

v5.13.0

NeuralPDE v5.13.0

Diff since v5.12.0

Merged pull requests:

• build(deps): bump codecov/codecov-action from 3 to 4 (#800) (@dependabot[bot])
• add Deep Galerkin method (#802) (@ayushinav)
• fix: bounds of variables in neural adapter setup (#812) (@sathvikbhagavan)
• CompatHelper: bump compat for DomainSets to 0.7, (keep existing compat) (#813) (@github-actions[bot])
• CompatHelper: bump compat for DomainSets to 0.7 for package docs, (keep existing compat) (#814) (@github-actions[bot])
• ci: explicitly specify token for codecov (#819) (@sathvikbhagavan)
• refactor: use latest API for IntegralProblem (#822) (@sathvikbhagavan)
• build(deps): bump crate-ci/typos from 1.18.2 to 1.19.0 (#823) (@dependabot[bot])
• chore: format docstrings of DGM to make it consistent with others (#825) (@sathvikbhagavan)
• reducing test times for DGM (#826) (@ayushinav)

Closed issues:

• Deep learning methods for free-boundary PDEs (#19)
• Deep Galerkin Method (#220)
• DGM Tutorial fails (#827)

v5.12.0

NeuralPDE v5.12.0

Diff since v5.11.0

Merged pull requests:

• Bump AdvancedHMC (#804) (@ChrisRackauckas)
• build(deps): bump crate-ci/typos from 1.18.0 to 1.18.2 (#805) (@dependabot[bot])
• fix: tstops loss in NNODE (#807) (@sathvikbhagavan)
• refactor: loss functions for NNODE (#808) (@sathvikbhagavan)
• build: bump Optimization packages and use only latest versions for all (#809) (@sathvikbhagavan)

Closed issues:

• NN ODE loss gets squared (#606)
• Problem in replicating INVERSE PROBELM tutorial on ODE (#765)
• Bump CUDA Compat (#810)

v5.11.0

NeuralPDE v5.11.0

Diff since v5.10.0

Merged pull requests:

• Bump actions/cache from 3 to 4 (#787) (@dependabot[bot])
• Remove duplicated methods in BPINN as it causes methods overwrite (#788) (@sathvikbhagavan)
• Remove Flux Support (#789) (@sathvikbhagavan)
• DAE problem (#790) (@KirillZubov)
• feat: add Parameter estimation capability in NNODE (#795) (@sathvikbhagavan)
• Add QA (#796) (@sathvikbhagavan)
• ci: add spell check workflow (#797) (@sathvikbhagavan)
• build(deps): bump crate-ci/typos from 1.17.0 to 1.18.0 (#801) (@dependabot[bot])

Closed issues:

• Recreate the examples from the BINN paper with NNODE (#640)
• Issue with NNStopping Example code (#680)
• NeuralPDE v.5.9 stops QuasiMonteCarlo from updating (#748)
• Parameter Estimation in NNODE (#766)

v5.10.0

NeuralPDE v5.10.0

Diff since v5.9.0

Merged pull requests:

• Docs modifications to use QuadratureTraining instead of GridTraining (#729) (@sdesai1287)
• Documenter 1.0 upgrade (#739) (@ArnoStrouwen)
• changes from final reviews, improved Docs (#741) (@AstitvaAggarwal)
• Remove v1.6 from CI (#743) (@ChrisRackauckas)
• BPINN PDE solver (#745) (@AstitvaAggarwal)
• ci: run Documentation builds on buildkite instead of GHA (#750) (@thazhemadam)
• build: make optim version >= 1.7.8 (#753) (@sathvikbhagavan)
• ci(buildkite): fix GPU tests not being run on Buildkite (#756) (@thazhemadam)
• Refactor/cleanup ode tutorials (#757) (@sathvikbhagavan)
• BNNODE bug fix, faster MvNormal pdf definition (#760) (@AstitvaAggarwal)
• Fix typo "simultaniously" -> "simultaneously" (#763) (@brollb)
• CompatHelper: add new compat entry for Statistics at version 1, (keep existing compat) (#764) (@github-actions[bot])
• Change typeof(x) <: y to x isa y (#767) (@pepijndevos)
• Fix ODE tutorial link (#771) (@bowenszhu)
• CompatHelper: bump compat for Adapt to 4, (keep existing compat) (#775) (@github-actions[bot])
• version 1.9 to dodge compat bump bandwagon (#776) (@AstitvaAggarwal)
• BPINN solver Docs(Manual and tutorial) (#781) (@AstitvaAggarwal)
• build: compat bumps to latest versions (#783) (@sathvikbhagavan)
• build: remove >= for Optim compat. (#786) (@sathvikbhagavan)

Closed issues:

• Migrate functionality to HighDimPDE.jl (#514)
• NNODE fails on simple logistic curve (#634)
• Bayesian Inference for PDE PINN solver (#712)
• fix NUTS kwargs handling and Default ode params vector (#737)
• BNNODE dosent support Lux chains Initparams, supporting fix for PR #749 (#759)
• Documentation for BayesianPINN and ahmc_bayesian_pinn_pde (#779)
• Encountering Precompilation Error Messages When Adding NeuralPDE Package in Julia (#780)