v0.2.4

ProbNumDiffEq v0.2.4

Diff since v0.2.3

Merged pull requests:

• Add compatibility to DiffEqDevTools.jl (#73) (@nathanaelbosch)

v0.2.3

ProbNumDiffEq v0.2.3

Diff since v0.2.2

Merged pull requests:

• Format the whole codebase with JuliaFormatter.jl (#71) (@nathanaelbosch)
• Improve the TaylorMode Initialization for SecondOrderODEProblems (#72) (@nathanaelbosch)

v0.2.2

ProbNumDiffEq v0.2.2

Diff since v0.2.1

Merged pull requests:

• More code optimization, with a big focus on the smoothing (#70) (@nathanaelbosch)

v0.2.1

ProbNumDiffEq v0.2.1

Diff since v0.2.0

Merged pull requests:

• Make the precompilation a tiny bit nicer (precompile when using) (#69) (@nathanaelbosch)

v0.2.0

ProbNumDiffEq v0.2.0

Diff since v0.1.9

Merged pull requests:

• More efficient initialization via TaylorIntegration.jl (#66) (@nathanaelbosch)
• New ordering of the states; New minor version! (#67) (@nathanaelbosch)
• Small optimizations throughouth the codebase (#68) (@nathanaelbosch)

v0.1.9

ProbNumDiffEq v0.1.9

Diff since v0.1.8

Merged pull requests:

• Support for OOP SecondOrderODEProblems (#54) (@nathanaelbosch)
• Better, more thorough convergence tests (#55) (@nathanaelbosch)
• Matrix-valued problems should work now (#56) (@nathanaelbosch)
• Support for scalar-valued ODE problems (#57) (@nathanaelbosch)
• Bugfix matrix-valued ODEs and remove support for scalar problems again (#58) (@nathanaelbosch)
• Only perform an update step if the step has not been rejected! (#59) (@nathanaelbosch)
• Renamed and restructured "measure!", to compute S separately (#60) (@nathanaelbosch)
• No more preconditioning in the forward-pass of the solver! (#61) (@nathanaelbosch)
• Only use Octavian.jl's matmul for non-abstract arrays (#62) (@nathanaelbosch)
• Made a first benchmark to compare ProbNumDiffEq.jl to various other solver packages (#63) (@nathanaelbosch)
• Initialization via non-probabilistic solvers (#64) (@nathanaelbosch)

v0.1.8

ProbNumDiffEq v0.1.8

Diff since v0.1.7

Merged pull requests:

• CompatHelper: bump compat for "ModelingToolkit" to "6" (#46) (@github-actions[bot])
• Better smoothing stability (#52) (@nathanaelbosch)
• Better preconditioning (#53) (@nathanaelbosch)

v0.1.7

ProbNumDiffEq v0.1.7

Diff since v0.1.6

Merged pull requests:

• The smoothing should go until the first index (#43) (@nathanaelbosch)
• Improvements to the README and documentation (#44) (@nathanaelbosch)
• Better error/warning message for dense!=alg.smooth (#45) (@nathanaelbosch)
• Simple change that removes the use of Taylor for q=1 initialization (#47) (@nathanaelbosch)
• Reduce memory allocations (#48) (@nathanaelbosch)
• Speed up the solvers with Octavian.jl (#49) (@nathanaelbosch)
• CompatHelper: add new compat entry for "Octavian" at version "0.3" (#50) (@github-actions[bot])
• CompatHelper: add new compat entry for "Tullio" at version "0.3" (#51) (@github-actions[bot])

v0.1.6

ProbNumDiffEq v0.1.6

Diff since v0.1.5

Merged pull requests:

• CompatHelper: bump compat for "StructArrays" to "0.6" (#37) (@github-actions[bot])
• Add SecondOrderODEProblem support (only IIP for the time being) (#40) (@nathanaelbosch)
• Added tests for automatic differentiation (#41) (@nathanaelbosch)
• Fix an issue with the default GR build in the documentation (#42) (@nathanaelbosch)

v0.1.5

ProbNumDiffEq v0.1.5

Diff since v0.1.4

Merged pull requests:

• Remove Julia 1.6 from testing again (#32) (@nathanaelbosch)
• Added ForwardDiff.jl as a dependency (#33) (@nathanaelbosch)
• CompatHelper: bump compat for "TaylorSeries" to "0.11" (#34) (@github-actions[bot])
• Use the proper Big types in the IBM prior (#35) (@nathanaelbosch)
• Made the state-initialization a bit easier (#36) (@nathanaelbosch)
• ForwardDiff for jacobian computation (#38) (@nathanaelbosch)
• Improved type stability in the projection matrices and in the preconditioner (#39) (@nathanaelbosch)