Carlos Parada

Results 443 comments of Carlos Parada

> After applying [this PR](https://github.com/JuliaInterop/RCall.jl/pull/496) all that's needed is to set things up in LocalPreference.toml in the appropriate way. Here is a script to do it automatically: > > ```...

> Hi @c42f, is this issue fixed? I am a new contributor and want to work on it ^^' Nope, but this looks like a great first issue! Feel free...

> Packages that offer partial eigendecompositions: > > * [Arpack.jl](https://github.com/JuliaLinearAlgebra/Arpack.jl) > * [ArnoldiMethod.jl](https://github.com/JuliaLinearAlgebra/ArnoldiMethod.jl) > * [KrylovKit.jl](https://github.com/Jutho/KrylovKit.jl) > * [IterativeSolvers.jl](https://github.com/JuliaLinearAlgebra/IterativeSolvers.jl) > > I am not sure that such iterative methods fit...

> > The power method seems basic, small, and well-established > > Also mostly obsolete. Krylov methods are generally strictly better, even with restarting. Huh, that's surprising, I thought the...

> If deemed sufficient, a basic version of the power algorithm can be easily implemented by users themselves. Definitely can be, but didn't @stevengj just say it's obsolete? The largest...

Yeah, if it's only really useful for educational reasons, it should go in a package. But the largest eigenvalue problem is common enough that I'd guess there should be a...

> Enabling shortcuts such as cmd S to save a file in vim could be a game changer for terminal usage. Things don't have to be cryptic and unfamiliar to...

> Then the problem for sandboxed applications would still remain. The regular user would not know or care for the difference between installing it from Flathub or from the distro's...

Agreed with basically all of these comments/changes, or else I don't disagree strongly enough to object. > * `ParetoSmooth.ess_sup`: While the one in `PSIS.ess_is` is a variance-based (i.e. l2-norm) ESS...

Generally, my recommendation is to go to Arviz.jl for LOO-related inference, as it provides more features. Otherwise, the issue here is closely related to #67: the problem is if we're...