Sourced from ModelingToolkit's changelog.

ModelingToolkit v11 Release Notes

Symbolics@7 and SymbolicUtils@4 compatibility

SymbolicUtils version 4 involved a major overhaul of the core symbolic infrastructure, which propagated to Symbolics as Symbolics version 7. ModelingToolkit has now updated to these versions. This includes significant type-stability improvements, enabling precompilation of large parts of the symbolic infrastructure and faster TTFX. It is highly recommended to read the Release Notes for SymbolicUtils@4 and the doc page describing the new variant structure before these release notes.

As part of these changes, ModelingToolkit has changed how some data is represented to allow precompilation. Notably, variable => value mappings (such as guesses) are stored as an AbstractDict{SymbolicT, SymbolicT}. Here, SymbolicT is a type that comes from Symbolics.jl, and is the type for all unwrapped symbolic values. This means that any non-symbolic values are stored as SymbolicUtils.Const variants. Mutation such as guesses(sys)[x] = 1.0 is still possible, and values are automatically converted. However, obtaining the value back requires usage of SymbolicUtils.unwrap_const or Symbolics.value.

Following is a before/after comparison of the TTFX for the most common operations in ModelingToolkit.jl. Further improvements are ongoing. Note that the timings do depend on many factors such as the exact system used, types passed to constructor functions, other packages currently loaded in the session, presence of array variables/equations, whether index reduction is required, and the behavior of various passes in mtkcompile. However, the numbers are good representations of the kinds of performance improvements that are possible due to the new infrastructure. There will continue to be improvements as this gets more extensive testing and we are better able to identify bottlenecks in compilation.

System constructor

The time to call System, not including the time taken for @variables or building the equations.

Before:

  0.243758 seconds (563.80 k allocations: 30.613 MiB, 99.48% compilation time: 3% of which was recompilation)
elapsed time (ns):  2.43757958e8
gc time (ns):       0
bytes allocated:    32099616
pool allocs:        563137
non-pool GC allocs: 16
malloc() calls:     651
free() calls:       0
minor collections:  0
full collections:   0

After:

</tr></table> 

... (truncated)

• See full diff in compare view

Sourced from ModelingToolkit's changelog.

ModelingToolkit v11 Release Notes

Symbolics@7 and SymbolicUtils@4 compatibility

SymbolicUtils version 4 involved a major overhaul of the core symbolic infrastructure, which propagated to Symbolics as Symbolics version 7. ModelingToolkit has now updated to these versions. This includes significant type-stability improvements, enabling precompilation of large parts of the symbolic infrastructure and faster TTFX. It is highly recommended to read the Release Notes for SymbolicUtils@4 and the doc page describing the new variant structure before these release notes.

As part of these changes, ModelingToolkit has changed how some data is represented to allow precompilation. Notably, variable => value mappings (such as guesses) are stored as an AbstractDict{SymbolicT, SymbolicT}. Here, SymbolicT is a type that comes from Symbolics.jl, and is the type for all unwrapped symbolic values. This means that any non-symbolic values are stored as SymbolicUtils.Const variants. Mutation such as guesses(sys)[x] = 1.0 is still possible, and values are automatically converted. However, obtaining the value back requires usage of SymbolicUtils.unwrap_const or Symbolics.value.

Following is a before/after comparison of the TTFX for the most common operations in ModelingToolkit.jl. Further improvements are ongoing. Note that the timings do depend on many factors such as the exact system used, types passed to constructor functions, other packages currently loaded in the session, presence of array variables/equations, whether index reduction is required, and the behavior of various passes in mtkcompile. However, the numbers are good representations of the kinds of performance improvements that are possible due to the new infrastructure. There will continue to be improvements as this gets more extensive testing and we are better able to identify bottlenecks in compilation.

System constructor

The time to call System, not including the time taken for @variables or building the equations.

Before:

  0.243758 seconds (563.80 k allocations: 30.613 MiB, 99.48% compilation time: 3% of which was recompilation)
elapsed time (ns):  2.43757958e8
gc time (ns):       0
bytes allocated:    32099616
pool allocs:        563137
non-pool GC allocs: 16
malloc() calls:     651
free() calls:       0
minor collections:  0
full collections:   0

After:

</tr></table> 

... (truncated)

• See full diff in compare view

Sourced from ForwardDiff's releases.

v1.4.1

ForwardDiff v1.4.1

Diff since v1.4.0

Merged pull requests:

• Remove allocations of nested Jacobians of StaticArrays (#810) (@​albert-de-montserrat)
• Bump codecov/codecov-action from 6 to 7 (#814) (@​dependabot[bot])

• 777420f Remove allocations of nested Jacobians of StaticArrays (#810)
• 5bf530a Bump codecov/codecov-action from 6 to 7 (#814)
• 0a3d297 CompatHelper: bump compat for LogExpFunctions to 1, (keep existing compat) (#...
• 39916da CompatHelper: bump compat for JET in [extras] to 0.11, (keep existing compat)...
• 7e52ffa Bump julia-actions/setup-julia from 2 to 3 (#802)
• 39a27fe Bump codecov/codecov-action from 5 to 6 (#799)
• 1295777 Avoid reshape allocation in extract_jacobian! for Matrix results (#797)
• ff0d903 Bump julia-actions/cache from 2 to 3 (#796)
• 7262054 Remove explicit != methods for Dual (#793)
• 71258ec Update documentation formatting and infrastructure (#792)
• Additional commits viewable in compare view

Sourced from Sundials's releases.

v6.2.1

Sundials v6.2.1

Diff since v6.2.0

Merged pull requests:

• Bump to SciMLLogging 2.0 (#529) (@​jClugstor)
• Bump julia-actions/setup-julia from 2 to 3 (#532) (@​dependabot[bot])

Sourced from Sundials's changelog.

Sundials.jl NEWS

v5.0.0

Breaking Changes

Upgrade to Sundials v7

This release updates the underlying Sundials C library from v5 to v7, which introduces significant API changes. This is a breaking change for users directly using the low-level Sundials API.

Key Changes:

1. SUNContext requirement: All Sundials objects now require a SUNContext object for creation. This context manages the Sundials environment and must be created before any solver objects.

2. Memory management: The new context-based approach improves thread safety and resource management.

Migration Guide for Low-Level API Users:

If you're using the low-level Sundials API directly (not through the DiffEq interface):

# Old code (v4.x) - No context needed
mem_ptr = CVodeCreate(CV_BDF)
mem = Handle(mem_ptr)

# New code (v5.0) - Context required
ctx_ptr = Ref{SUNContext}(C_NULL)
SUNContext_Create(C_NULL, Base.unsafe_convert(Ptr{SUNContext}, ctx_ptr))
ctx = ctx_ptr[]
mem_ptr = CVodeCreate(CV_BDF, ctx)  # Context passed as argument
mem = Handle(mem_ptr)
# ... use solver ...
SUNContext_Free(ctx)  # Clean up context when done

Automatic handling in high-level interface:

If you're using the standard DiffEq interface (solve(prob, CVODE_BDF())), no changes are needed. The context is automatically managed internally:

# This continues to work without changes
sol = solve(prob, CVODE_BDF())

Functions affected by context requirement:

• All solver creation functions (CVodeCreate, ARKStepCreate, IDACreate, KINCreate)
• All vector creation functions (N_VNew_Serial, etc.)
• All matrix creation functions (SUNDenseMatrix, SUNBandMatrix, etc.)

... (truncated)

• 1ab4168 Bump to SciMLLogging 2.0 (#529)
• e174eef Merge pull request #532 from SciML/dependabot/github_actions/julia-actions/se...
• a520482 Bump julia-actions/setup-julia from 2 to 3
• 32c446a Update Project.toml
• 70cb9d4 Merge pull request #531 from ChrisRackauckas-Claude/fix-master-derivative-dis...
• 758a77f ci: bump Sundials compat in test/jet/Project.toml to include v6
• bb43976 test: port bouncing-ball VectorContinuousCallback test to DiffEqBase v7 API
• d5494d2 Apply suggestion from @​ChrisRackauckas
• 70f6af9 Forward derivative_discontinuity getproperty/setproperty! to u_modified
• 8bb8018 Merge pull request #530 from ChrisRackauckas-Claude/fix-formatcheck-julia-setup
• Additional commits viewable in compare view

Sourced from Symbolics's releases.

v7.28.1

Symbolics v7.28.1

Diff since v7.28.0

Merged pull requests:

• add fastpaths for zero/one for Num (#1902) (@​KristofferC)
• _depends_on: avoid looking up COMPARE_FULL in every iteration and directly call isequal_bsimpl (#1903) (@​KristofferC)
• build: bump patch version (#1904) (@​AayushSabharwal)

Sourced from Symbolics's changelog.

7.0.0

Breaking changes

• substitute no longer recurses into Differential arguments. This is due to SymbolicUtils.jl v4's default_substitute_filter, which treats Operator subclasses (including Differential) as substitution boundaries. Use the new substitute_in_deriv or substitute_in_deriv_and_depvar functions to substitute inside Differential expressions. See the Derivatives documentation for details.

4.0.0

• Symbolics.jl now supports the latest symbolic computing architecture backed by Metatheory.jl v1.2 and SymbolicUtils.jl v0.18 for generic term rewriting.
• Support for automatic code optimization through Metatheory.jl EGraphs and SymbolicUtils's optimize function.

3.3.0

• adds simplify_fractions which turns an expression into a single fraction and simplifies by dividing the numerator and denominator factors by appropriate GCDs
• Use new fraction_iszero and fraction_isone functions from SymbolicUtils to implement iszero and isone respectively.
• x / x etc. are no more simplified on construction, call simplify_fractions to simplify them.

• See full diff in compare view

Sourced from OrdinaryDiffEq's changelog.

OrdinaryDiffEq.jl v7 / DifferentialEquations.jl v8 Breaking Changes

This release bumps to SciMLBase v3, RecursiveArrayTools v4, and includes breaking changes across DiffEqBase, OrdinaryDiffEqCore, and all solver sublibraries. It also coincides with the DifferentialEquations.jl v8 umbrella release, which is itself a breaking change to the user-facing meta-package.

DifferentialEquations.jl v8: scope reduction

DifferentialEquations.jl v8 no longer re-exports the full SciML solver suite. Previously, using DifferentialEquations pulled in OrdinaryDiffEq, StochasticDiffEq, DelayDiffEq, BoundaryValueDiffEq, Sundials, JumpProcesses, SteadyStateDiffEq, LinearSolve, NonlinearSolve, Optimization, etc. — a large default surface that drove up using time and made it unclear which package any given solver actually came from.

In v8, using DifferentialEquations only loads OrdinaryDiffEq. All other solver families have been removed from the umbrella. If your code relied on DifferentialEquations for SDEs, DDEs, BVPs, jumps, steady states, or any non-ODE solver, you will need to add the topic-specific package to your project explicitly.

Migration

Find the topic you need a solver for and add the corresponding sublib(s) directly. The DiffEqDocs tutorials and solver pages now specify, per algorithm, which package it ships from. Common cases:

Topic	Old (DiffEq v7 umbrella)	New (DiffEq v8)
ODEs	using DifferentialEquations	using OrdinaryDiffEq (or using OrdinaryDiffEqTsit5, OrdinaryDiffEqRosenbrock, … for individual solver families)
Stochastic ODEs	using DifferentialEquations	using StochasticDiffEq
Delay ODEs	using DifferentialEquations	using DelayDiffEq
Boundary value problems	using DifferentialEquations	using BoundaryValueDiffEq (or one of BoundaryValueDiffEqMIRK, BoundaryValueDiffEqFIRK, BoundaryValueDiffEqShooting, …)
Jump processes	using DifferentialEquations	using JumpProcesses
Steady state	using DifferentialEquations	using SteadyStateDiffEq
DAEs (mass matrix or implicit)	using DifferentialEquations	using OrdinaryDiffEq (mass matrix), using Sundials (IDA), or topic sublib
Sundials wrappers (CVODE, IDA, ARKODE)	using DifferentialEquations	using Sundials
Linear / nonlinear / optimization	using DifferentialEquations	using LinearSolve / using NonlinearSolve / using Optimization

For ODE work specifically, prefer importing only the sublib you need (e.g. using OrdinaryDiffEqTsit5: Tsit5) rather than the umbrella using OrdinaryDiffEq — the v7 ecosystem split lets you trim using time substantially. The DiffEqDocs tutorials and solver index annotate every algorithm with its host sublib.

Why

Removing the meta-package's broad re-exports lets each topic's package version cycle independently, eliminates the long using DifferentialEquations precompile chain for users who only need ODEs, and makes the dependency graph for any given script honest about what's actually being loaded.

This change is independent of the OrdinaryDiffEq v7 changes below — OrdinaryDiffEq v7 ships with DifferentialEquations v8, but you can also use OrdinaryDiffEq v7 directly without the umbrella package at all.

OrdinaryDiffEq.jl v7 Breaking Changes

This release bumps to SciMLBase v3, RecursiveArrayTools v4, and includes breaking changes across DiffEqBase, OrdinaryDiffEqCore, and all solver sublibraries.

Themes of the v7 release

Most of the breaking changes fall into a small set of recurring themes. Keep these in mind while reading the migration table — they explain why an individual change exists and often suggest the right migration direction:

• Time to first solve (TTFS) reduction. Direct deps on Static.jl, StaticArrayInterface.jl, Polyester.jl, and StaticArrays.jl were dropped; using OrdinaryDiffEq now loads only the default solver set; ODEFunction switched to AutoSpecialize. All of this means less code loaded and more precompilation caching on first solve.

• Type stability everywhere. All Bool solver/solve keyword arguments (autodiff, verbose, alias, lazy, …) were replaced by typed objects. Passing a Bool no longer changes dispatch in ways the compiler cannot specialize on, and the reverse is no longer allowed to silently fall back through slow generic paths. For example:

  # v6 — a Bool
  Rosenbrock23(autodiff = true)
  solve(prob, alg; verbose = false, alias = true)

... (truncated)

• e288f7f Update Project.toml
• 48e9405 export scimllogging (#3766)
• 0bda310 Change typo in NEWS (#3761)
• 84716be Add StochasticDiffEq docs to lib/StochasticDiffEq/docs (#3759)
• 8d56670 Merge pull request #3760 from oscardssmith/os/bump-OrdinaryDiffEq-version
• 6c7cbfe bump version
• c538565SciML/.github#77#3738)
• 0df6173 Document EulerHeun as two-stage Stratonovich Heun method (#3167) (#3757)
• bce6d6c Downgrade: raise RecursiveArrayTools compat floor to 4.2.0 (old absent floor ...
• 13e7f0f Use SciMLTesting v1.2 folder-based run_tests (#3752)
• Additional commits viewable in compare view

Sourced from DataInterpolations's releases.

v8.10.0

DataInterpolations v8.10.0

Diff since v8.9.0

Merged pull requests:

• build(deps): bump JuliaRegistries/TagBot from 1.24.2 to 1.24.3 (#501) (@​dependabot[bot])
• Create clarification.md for interpolation concepts (#502) (@​ajatshatru01)
• Fix Runic formatting in docs/make.jl (#503) (@​ChrisRackauckas-Claude)
• Fix issorted check failing with ForwardDiff.Dual for duplicate time points (#513) (@​ChrisRackauckas-Claude)
• Mooncake AD for DataInterpolations (#514) (@​AstitvaAggarwal)

Closed issues:

• Failure to construct interpolation during ForwardDiff gradient calculation with duplicated time point (#510)

Sourced from DataInterpolations's changelog.

DataInterpolations v9 Release Notes

Breaking changes

• The deprecated RegularizationTools extension and the RegularizationSmooth interpolation type have been removed. RegularizationTools was deprecated and capped Optim to ≤ 1; removing it restores support for Optim v2.

DataInterpolations v5 Release Notes

Breaking changes

• AbstractInterpolation is not a subtype of AbstractVector anymore. This was needed for previous versions of ModelingToolkit.jl to represent splines as vectors.

• Indexing overloads for AbstractInterpolation and the type parameter associated with it are removed. For example - A is an interpolation object:

  • Doing A[i] will error. Use A.u[i].
  • size(A) will error. Use size(A.u) or size(A.t).

• Removed deprecated bindings for ZeroSpline which is the same as ConstantInterpolation.

DataInterpolations v6 Release Notes

Breaking changes

• SciML/DataInterpolations.jl#274 introduced caching of parameters for interpolations (released in v5.3) and also introduced a field safetycopy which was a boolean flag to create a copy of the data as the parameters would be invalid if data is mutated. This was removed in SciML/DataInterpolations.jl#315 to introduce cache_parameters which made it explicit if a user wants to opt in for parameter caching or not.

• See full diff in compare view

Sourced from DifferentialEquations's releases.

v8.0.0

DifferentialEquations v8.0.0

Diff since v7.17.0

Many breaking changes. The complete migration story is detailed in https://github.com/SciML/OrdinaryDiffEq.jl/blob/master/NEWS.md

Sourced from DifferentialEquations's changelog.

Breaking Changes in v7.0

• ParameterizedFunctions.jl, along with a few other modeling libraries, are no longer exported by DifferentialEquations.jl. You must do using ParameterizedFunctions to use the `@ode_def macro. As it has not been part of the documentation for 3 years, this is breaking but should not affect most users. This reduces the dependencies and using time of the library by about half.
• OrdinaryDiffEq.jl has a new linear solver system based on the LinearSolve.jl library. linsolve arguments now take LinearSolve.jl solvers.

• c560df7 Merge pull request #1137 from ChrisRackauckas-Claude/bump-ordinarydiffeq-v7-e...
• 4ae2312 Exclude SVG assets from typos spell check
• 2f0f3de Fix Runic formatting: use explicit float literals
• 33b3433 Bump compat for OrdinaryDiffEq v7 / SciMLBase v3 ecosystem
• bf443f2 Merge pull request #1135 from SciML/dependabot/github_actions/julia-actions/s...
• 4300c87 Bump julia-actions/setup-julia from 2 to 3
• bfd3144 Merge pull request #1126 from ChrisRackauckas-Claude/fix-deprecation-warnings
• 9188ef8 Add comprehensive test suite using non-deprecated APIs
• e0329c6 Merge pull request #1117 from SciML/dependabot/github_actions/actions/checkout-6
• f306ba6 Bump actions/checkout from 4 to 6
• Additional commits viewable in compare view

You can trigger Dependabot actions by commenting on this PR:

• @dependabot rebase will rebase this PR
• @dependabot recreate will recreate this PR, overwriting any edits that have been made to it
• @dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
• @dependabot ignore <dependency name> major version will close this group update PR and stop Dependabot creating any more for the specific dependency's major version (unless you unignore this specific dependency's major version or upgrade to it yourself)
• @dependabot ignore <dependency name> minor version will close this group update PR and stop Dependabot creating any more for the specific dependency's minor version (unless you unignore this specific dependency's minor version or upgrade to it yourself)
• @dependabot ignore <dependency name> will close this group update PR and stop Dependabot creating any more for the specific dependency (unless you unignore this specific dependency or upgrade to it yourself)
• @dependabot unignore <dependency name> will remove all of the ignore conditions of the specified dependency
• @dependabot unignore <dependency name> <ignore condition> will remove the ignore condition of the specified dependency and ignore conditions