Mathieu Besançon

weak separation oracle

2

introduces weak separation oracles as an abstract interface that other FW algorithms will be able to leverage. Right now the only weak oracles are cache-based (active set or cache) in...

Ordered Weighted Norm

2

Implement the LMO for ordered weighted norms https://arxiv.org/pdf/1409.4271.pdf

relaxation handler

4

expose a relaxation handler in the wrapper with an idiomatic interface

Inconsistent addition between sparse and dense

1

```julia julia> spzeros(2) + [1.4, 3] 2-element SparseVector{Float64, Int64} with 2 stored entries: ⋮ [2] = 3.0 julia> spzeros(2,2) + [1.4 3 0 2] 2×2 Matrix{Float64}: 1.4 3.0 0.0 2.0...

```julia using SparseArrays M = spzeros(3,3) v = spzeros(3) copyto!(@view(M[:,2]), v) ``` calls the slow fallback `copyto!(dest::AbstractArray, src::AbstractArray)`. Possible fix would be to define copyto! with: ```julia copyto!(::SubArray{T, N,

Generalize strong convexity to sharpness following the paper

[JOSS] make functions run deterministic

23

Following the convention from Julia stdlib's Random and various packages from the JuliaStats ecosystem, every function which performs a random sampling should take in parameter the Random Number Generator (which...

treeplot function

5

```julia function treeplot(trr::Tree, fig= 1) if !isempty(fig) figure(fig) end # ... ``` it is not trivial what exactly users should pass as `fig` argument. `isempty` for a number is always...