Remove smoothing.

Changelog.md

@@ -11,8 +11,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 * the Levenberg-Marquardt algorithm internally uses a `VectorGradientFunction`, which allows
  to use a vector of gradients of a function returning all gradients as wel for the algorithm
-* smoothing is now available for the Levenberg-Marquardt algorithm
-
+* The `VectorGradientFunction` now also have a `get_jacobian` function
 
 ### Changed
 

docs/src/solvers/LevenbergMarquardt.md

@@ -15,12 +15,6 @@ LevenbergMarquardt!
 LevenbergMarquardtState
 ```
 
-## Available smoothing functions
-
-```@docs
-Manopt.smoothing_factory
-```
-
 ## [Technical details](@id sec-lm-technical-details)
 
 The [`LevenbergMarquardt`](@ref) solver requires the following functions of a manifold to be available

src/Manopt.jl

@@ -391,7 +391,6 @@ export get_state,
     forward_operator!,
     get_objective,
     get_unconstrained_objective
-export smoothing_factory
 export get_hessian, get_hessian!
 export ApproxHessianFiniteDifference
 export is_state_decorator, dispatch_state_decorator

src/documentation_glossary.jl

@@ -268,13 +268,11 @@ define!(
     :NonLinearLeastSquares,
     (; M="M", p="p") -> """
 ```math
-$(_tex(:argmin))_{$p ∈ $(_math(:M; M=M))} $(_tex(:frac,1,2)) $(_tex(:sum))_{i=1}^n s_i$(_tex(:bigl))($(_tex(:bigl))( $(_tex(:abs, "f_i($p)"))^2 $(_tex(:bigr)))
+$(_tex(:argmin))_{$p ∈ $(_math(:M; M=M))} $(_tex(:frac,1,2)) $(_tex(:sum))_{i=1}^n $(_tex(:abs, "f_i($p)"))^2
 ```
 
 where ``f: $(_math(:M; M=M)) → ℝ^n`` is written with component functions ``f_i: $(_math(:M; M=M)) → ℝ``,
 and each component function is continuously differentiable.
-The function ``s_i: ℝ → ℝ`` can be seen as smoothing or regularisation  of the least squares term.
-It is assumed to be twice continuously differentiable and its default is ``s_i(x) = x`` for all ``i=1,…n``.
 """,
 )