that will be useful for problems where we can take the concrete intersection between the union and the guard
a comment on LazySets and ModelingToolkit (cc @sebastianguadalupe) note that instead of
@variables x v
X = HPolyhedron([HalfSpace(x ≥ 0, vars),
HalfSpace(v ≤ 0, vars)])we can write
X = HPolyhedron([x ≥ 0, v ≤ 0], vars)
it would be nice that we can also parse double inequalities such as
0 <= x <= 1. but they are not yet available in ModelingToolkit (see SciML/ModelingToolkit.jl#589 )
i think though that we may be able to parse such expression with a macro along these lines
X = @set(x ≥ 0, 0 ≤ v ≤ 1, vars)
@SebastianGuadalupe have you tried if it works with the latest versions of the packages?
but we can change it to SXFormat.jl, for instance
or SXParser.jl
any opinions on this change?
SpaceExParser sounds good to me.. it's self-descriptive :)
but I didn't use it too much either
@SebastianGuadalupe ok. note that the build failure in JuliaReach/ReachabilityModels.jl#31 is because it's not registered
i mean we can tweak this with the travis script, but i dont see a reason not to register the package
( btw i have interest in reviving that project -- it hasnt received attention in a while)
You should come !
although with 2 small kids now my traveling pass is compromised :)
starting in 20 mins
how much can be said in 3 minutes :o
which one have you checked?
i have a 100+ dimension tensor, which is significantly sparse. with only a few million non empty elements, and i need to do a lot of contraction out of them
but in principle i would expect that packages implement the abstract array interface
in which case you can just drop in any array to work with, in particular those from SparseArrays