You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
ChainRulesCore.jl `rrule` for integration with automatic differentiation libraries.
Note that it gives the gradient only with respect to the evaluation point `x`,
and not the data inside `itp`.
"""
function ChainRulesCore.rrule(itp::AbstractInterpolation, x...)
y = itp(x...)
function interpolate_pullback(Δy)
nope = ChainRulesCore.@not_implemented "`Interpolations.gradient` does not calculate a gradient with respect to the original data, only the evaluation point"
(nope, (sum(Δy .* g) for g in Interpolations.gradient(itp, x...))...)