v7.71.2

SciMLSensitivity v7.71.2

Diff since v7.71.1

Merged pull requests:

• Don't allow picking GaussAdjoint with TrackerVJP in default alg (#1145) (@ChrisRackauckas)

Closed issues:

• Default algorithm should have checks for Adjoints limitations (#1116)

v7.71.1

SciMLSensitivity v7.71.1

Diff since v7.71.0

Merged pull requests:

• Fix function grabbing in forward sensitivity remake (#1144) (@ChrisRackauckas)

Closed issues:

• remake incorrect for ODEForwardSensitivityProblem on 7.71 (#1143)

v7.71.0

SciMLSensitivity v7.71.0

Diff since v7.70.0

Merged pull requests:

• chore: use tunables in vec_pjac (#1097) (@DhairyaLGandhi)
• Avoid needing adjoints of SciMLStructures' constructor (#1135) (@DhairyaLGandhi)
• Fix remake of ForwardDiffSensitivity (#1142) (@ChrisRackauckas)

Closed issues:

• better error message for 'wrong' tspan inputs (#1137)

v7.70.0

SciMLSensitivity v7.70.0

Diff since v7.69.0

Merged pull requests:

• Update docs remove extra returns from loss and extra args from callback (#1128) (@Vaibhavdixit02)
• CompatHelper: bump compat for GPUArraysCore to 0.2, (keep existing compat) (#1134) (@github-actions[bot])
• Set SciMLBase bound (#1138) (@ChrisRackauckas)
• CompatHelper: bump compat for Functors to 0.5, (keep existing compat) (#1140) (@github-actions[bot])

v7.69.0

SciMLSensitivity v7.69.0

Diff since v7.68.0

Merged pull requests:

• refactor: cleanup deps and replace SparseDiffTools with SciMLJacobianOperators (#1124) (@avik-pal)
• CompatHelper: bump compat for DiffEqCallbacks to 4 for package docs, (keep existing compat) (#1126) (@github-actions[bot])
• CompatHelper: bump compat for DiffEqCallbacks to 4 for package gpu, (keep existing compat) (#1127) (@github-actions[bot])

v7.68.0

SciMLSensitivity v7.68.0

Diff since v7.67.0

Merged pull requests:

• Reflect ODEForwardSensitivityProblem kwarg change in docs (#1112) (@jClugstor)
• CompatHelper: bump compat for DiffEqFlux to 4 for package gpu, (keep existing compat) (#1113) (@github-actions[bot])
• CompatHelper: bump compat for Enzyme to 0.13, (keep existing compat) (#1114) (@github-actions[bot])
• CompatHelper: bump compat for Enzyme to 0.13 for package docs, (keep existing compat) (#1115) (@github-actions[bot])
• CompatHelper: bump compat for Optimization to 4 for package docs, (keep existing compat) (#1117) (@github-actions[bot])
• CompatHelper: bump compat for OptimizationOptimisers to 0.3 for package docs, (keep existing compat) (#1118) (@github-actions[bot])
• CompatHelper: bump compat for OptimizationOptimJL to 0.4 for package docs, (keep existing compat) (#1119) (@github-actions[bot])
• CompatHelper: bump compat for OptimizationPolyalgorithms to 0.3 for package docs, (keep existing compat) (#1121) (@github-actions[bot])
• Update Project.toml (#1123) (@ChrisRackauckas)

Closed issues:

• Neural SDE example fails on convert(::Type{Float64}, x::Tracker.TrackedReal{Float64}) (#1109)

v7.67.0

SciMLSensitivity v7.67.0

Diff since v7.66.2

Merged pull requests:

• CompatHelper: bump compat for Lux to 1 for package docs, (keep existing compat) (#1108) (@github-actions[bot])
• Split QA tests to a separate group (#1111) (@ChrisRackauckas)

Closed issues:

• missing method matching recursive_copyto!(::AbstractVector, ::Tuple) (#1082)

v7.66.2

SciMLSensitivity v7.66.2

Diff since v7.66.1

Merged pull requests:

• Switch positional alg argument to kwarg sensealg for ODEForwardSensitivityProblem (#1106) (@jClugstor)

Closed issues:

• Consistency in sensealg arguments (#1088)

v7.66.1

SciMLSensitivity v7.66.1

Diff since v7.66.0

Merged pull requests:

• Add support for using non-AbstractArrays with OrdinaryDiffEq and Zygote (#1092) (@apkille)
• Remove Deps: StochasticDiffEq and OrdinaryDiffEq (#1096) (@prbzrg)

Closed issues:

• [Deprecation warnings] use Base.getindex(A, :, i) instead. & Change A[i] to A.u[i] (#960)
• Depending on OrdinaryDiffEq? (#1095)

v7.66.0

SciMLSensitivity v7.66.0

Diff since v7.65.0

Merged pull requests:

• Fake solver output type to please Enzyme (#1098) (@ChrisRackauckas)
• Fix null parameters with ForwardDiffSensitivity (#1104) (@ChrisRackauckas)