Christian Schilling
Christian Schilling
https://github.com/JuliaReach/LazySets.jl/blob/e4995d5e4cb91082463071b22f5ed1d058116fc6/src/Sets/Ball2.jl#L324-L326 https://github.com/JuliaReach/LazySets.jl/blob/e4995d5e4cb91082463071b22f5ed1d058116fc6/src/Utils/samples.jl#L55-L60 In particular add a new sampler (as the default for `Ball2`). It is probably better to have only one `sample` method, so we should change the current method...
[Sampling from the Complement of a Polyhedron: An MCMC Algorithm for Data Augmentation](https://rafidrm.github.io/papers/MCMC_Sampling_Complement.pdf)
[Shake-and-Bake Algorithms for Generating Uniform Points on the Boundary of Bounded Polyhedra](http://deepblue.lib.umich.edu/bitstream/2027.42/3700/5/bam3210.0001.001.pdf)
Algorithm: https://gis.stackexchange.com/a/22904 Implementation: https://gis.stackexchange.com/a/22934
```julia julia> H = [1.0 1.0; 0.0 1.0]; julia> o = [2.5, 2.0, -4.0, -1.0]; julia> Pl = HParallelotope(H, o); # this set is empty julia> vertices_list(Pl) 4-element Vector{Vector{Float64}}: [2.0,...
Follow-up of #1178. In iterative refinement, we compute the support *vectors*. However, I would say that we could instead compute the support *functions*, which can be more efficient for certain...
In #2849 we assumed that the sets in the union are bounded. This can be generalized to unbounded sets. One idea would be to do a preprocessing of the unbounded...
Closes #1493. ~Closes #1711.~ * [x] remove `MathProgInterface` and `GLPKMathProgInterface`; load `MathOptInterface` * [x] refactor LP creation in `remove_redundant_constraints!` (`AbstractPolyhedron_functions.jl`) (the most complicated place because LPs are created in a...
For other `AbstractDirections` this is more difficult.
Closes #202. I had to find out that Julia is not very efficient for optional arguments, even if the default argument is chosen (I had assumed that this case can...