DataInterpolations.jl and Zygote.jl

ChrisRackauckas · April 18, 2024, 7:18pm

It currently only has the adjoints to differentiate w.r.t. t

SciML/DataInterpolations.jl/blob/master/ext/DataInterpolationsChainRulesCoreExt.jl#L15-L26


      
          function ChainRulesCore.rrule(::typeof(_interpolate),
                  A::Union{
                      LagrangeInterpolation,
                      AkimaInterpolation,
                      BSplineInterpolation,
                      BSplineApprox
                  },
                  t::Number)
              deriv = derivative(A, t)
              interpolate_pullback(Δ) = (NoTangent(), NoTangent(), deriv * Δ)
              return _interpolate(A, t), interpolate_pullback
          end

Topic		Replies	Views
Optimization.jl, DataInterpolations.jl and Gradients General Usage interpolations , gradient , data-interpolations	4	810	April 21, 2023
Zygote Performance Machine Learning question	22	5035	September 23, 2019
[ANN] Yota.jl - yet another reverse-mode autodiff package Package Announcements	28	3823	March 15, 2019
Is this impossible for Zygote Adjoint to handle General Usage zygote	4	613	April 7, 2021
Help optimizing 1000x Zygote overhead for linear interpolation General Usage	3	1665	April 15, 2022

DataInterpolations.jl and Zygote.jl

Related topics