Moritz Schauer comments

Results 302 comments of


                                            Moritz Schauer

use deterministic RNG initial state

Ping @haampie

use deterministic RNG initial state

@stevengj What to do with the tests? I agree that we can soften a bound for the second failure.

use deterministic RNG initial state

There is one test passing, > CI / Julia 1.6 - ubuntu-latest - x64 - pull_request (pull_request) which seems to be that the lobpg issue is fixed on some level....

Reference for the bivariate CDF approximation

I think what we really should do is calling https://github.com/JuliaStats/StatsFuns.jl/blob/e21bc26b1773aeb86bc8df4832db9c371d09ba2b/src/tvpack.jl#L338 which was added in the mean time and is a proper implementation for sure.

Merge collect() into Vector()?

`collect` returns an Array, not (always) a Vector. To use`collect` rather than `convert(Array,...)` indicates better that values/objects ) are to be gathered which possibly do not yet exists and which...

Merge collect() into Vector()?

By the way `convert(Array, 1:3)` is defined anyway because `isa(1:3,AbstractArray)` is true.

Merge collect() into Vector()?

With that, `Array` can do what `collect` currently does. Is the plan to deprecate `collect`? Just that `collect` has the ability to choose the right container for an iterator, which...

Merge collect() into Vector()?

... if you created `itr` yourself, of course. But Julia has a rudimentary system in place where iterators inform about the mutable collection they are naturally collected into via traits...

AbstractRNG Interface

Do you need a particular sequence such as golden integers or floats or golden tuples of those?

Sampling from a dirac delta

Related to e.g. #27249. There is `check=false` we can try, but maybe this gives garbage instead of positive semidefinite factors. In the end maybe we have to diverge from Base...