Christian Schilling
Christian Schilling
We have the function `tune_δ`. It would be nice if this function could receive more parameters for tuning. For instance in the `TMJets` algorithm we would also adapt `abs_tolerance`.
See [this notebook](https://github.com/mforets/escritoire/blob/master/reachability/BenchsHybrid.ipynb).
Same idea as in #577, but for the thermostat model we get this error in `constraints_list` of `Zonotope`: `LoadError: 1D-zonotope constraints currently only support a single generator`
It is probably enough to check that the `hcat` of all blocks is `1:n`. related: #92
What happens for the set at position `N`? _Originally posted by @schillic in https://github.com/JuliaReach/Reachability.jl/pull/537_
`merge(𝑂::Options, 2L𝑂::TwoLayerOptions) = merge(2LO.defaults, 𝑂, 2LO.specified)`
Currently every discrete-post operator wraps the reach sets in `ReachSet` objects in the `tube⋂inv!` method. Instead, this should be done at a common place (probably `solve` right before or after...
We had the option `:N`, but it got removed in #464.
I suggest to remove the special handling of `N == 1`. This error-prone code is a minor optimization for a case that never occurs.