v7.2.0

DataInterpolations v7.2.0

Diff since v7.1.0

Merged pull requests:

• Add Unitful tests for constant extrapolation (#381) (@Ickaser)
• Fix integrating over interval that is fully outside data range (#386) (@SouthEndMusic)
• Fix type inference and performance problems of munge_data (#392) (@devmotion)
• Improve cumulative_integral (#393) (@devmotion)
• Do not fall back to inplace interpolation method (#394) (@devmotion)

Closed issues:

• Broken extrapolation with integer interpolation (#378)
• AkimaInterpolation of Float32 data results in Float64 values when interpolating (#382)
• Integrals with incorrect dimensions for Unitful input (#385)
• Inconsistent output eltype when evaluating on vectors (#388)
• Regressions of ConstantInterpolation reported by JET.@report_opt (#389)

v7.1.0

DataInterpolations v7.1.0

Diff since v7.0.1

Merged pull requests:

• Specialize ConstantInterpolation extrapolation (address #378) (#379) (@DaniGlez)
• refactor: zygote rules for matrix inputs (#380) (@sathvikbhagavan)
• Remove hardcoded Float64s in AkimaInterpolation (#383) (@ChrisRackauckas)
• Bump Zygote compat to 0.7 (#384) (@SebastianM-C)

Closed issues:

• Does not support gradient calculation of 2D interpolation results (#371)

v7.0.1

DataInterpolations v7.0.1

Diff since v7.0.0

Merged pull requests:

• CompatHelper: bump compat for "DataInterpolations" to "7.0" for package docs (#374) (@github-actions[bot])
• Correct typo (#375) (@hersle)
• Fix ConstantInterpolation integer extrapolation (#376) (@SouthEndMusic)
• Bump Project.toml to 7.0.1 (#377) (@DaniGlez)

Closed issues:

• 1-point ConstantInterpolation gets BoundsError with cache_parameters (#373)

v7.0.0

DataInterpolations v7.0.0

Diff since v6.6.0

Breaking changes

• extrapolate keyword argument was removed. A new keyword argument extrapolation was added. The reason for this is the previous argument was a Boolean, whether to extrapolate or not, while the new argument is an extrapolation choice, i.e. how to extrapolate, like constant extrapolation, linear, continuing the polynomial, etc. See the docs for the choices.

Merged pull requests:

• Extrapolation options (#356) (@SouthEndMusic)
• Refactor integration and QuadraticInterpolation (#359) (@SouthEndMusic)
• build(deps): bump julia-actions/setup-julia from 2.6.0 to 2.6.1 (#362) (@dependabot[bot])
• Add extrapolation types periodic and reflective (#363) (@SouthEndMusic)
• Make extrapolation with NaN Unitful-friendly (#366) (@Ickaser)
• Fix merge issues (#369) (@SouthEndMusic)
• build(deps): bump JuliaRegistries/TagBot from 1.20.0 to 1.20.1 (#372) (@dependabot[bot])

Closed issues:

• Configurable extrapolation behavior (#355)
• Add extrapolation types periodic and reflection (#360)
• Error rather than NaN result if Unitful arrays passed to LinearInterpolation (#365)

v6.6.0

DataInterpolations v6.6.0

Diff since v6.5.2

Merged pull requests:

• ci: test with 1, lts and pre versions of julia (#347) (@thazhemadam)
• BSplineInterpolation: Add support for higher order arrays (#349) (@ashutosh-b-b)
• build(deps): bump JuliaRegistries/TagBot from 1.19.6 to 1.20.0 (#352) (@dependabot[bot])
• Update QuadraticSpline (#354) (@SouthEndMusic)

Closed issues:

• Unstable quadratic spline (#353)

v6.5.2

DataInterpolations v6.5.2

Diff since v6.5.1

Merged pull requests:

• build(deps): bump julia-actions/setup-julia from 2.5.0 to 2.6.0 (#345) (@dependabot[bot])
• Intercept frule at the top level (#346) (@YingboMa)

v6.5.1

DataInterpolations v6.5.1

Diff since v6.5.0

Merged pull requests:

• CubicSplines: Add AbstractMatrix support (#343) (@ashutosh-b-b)

v6.5.0

DataInterpolations v6.5.0

Diff since v6.4.5

Merged pull requests:

• refactor: add output dimensionality as a parameter in AbstractInterpolations (#342) (@sathvikbhagavan)

Closed issues:

• Indicate output dimensionality in AbstractInterpolation type (#328)

v6.4.5

DataInterpolations v6.4.5

Diff since v6.4.4

Merged pull requests:

• fix: zero(Vector(Vector)) doesn't work in 1.10 (#341) (@avik-pal)

v6.4.4

DataInterpolations v6.4.4

Diff since v6.4.3

Merged pull requests:

• fix: non-scalar tangents (#340) (@avik-pal)