Generalizing minimal initializers

[MartinuzziFrancesco]

Checklist

• Appropriate tests were added
• Any code changes were done in a way that does not break public API
• All documentation related to code changes were updated
• The new code follows the
  contributor guidelines, in particular the SciML Style Guide and
  COLPRAC.
• Any new documentation only uses public API

Additional context

Closes #265. I am thinking if the building pieces need to be user facing too. I need to update existing documentation to reflect the new flexibility of the initializers (option of using vector of values for weight). Add docstrings to the building pieces too if they become public api.

[MartinuzziFrancesco]

I think that we can generalize the minimal reservoir inits even more now by also having some random sign changes there. For this we need to format irrational and bernoulli to simply modify the arrays/matrices in place. Something like

function bernoulli!(rng::AbstractRNG, vecormat::AbstractVecOrMat; p::Number=0.5)
    for idx in eachindex(vecormat)
        if rand(rng) > p # keeping current approach consistent 
            vecormat[idx] = - vecormat[idx]
        end
    end
end

and then having the option to change signs in the inits functions like so

function delay_line!(rng::AbstractRNG, reservoir_matrix::AbstractMatrix, weight::AbstractVector,
        shift::Int; sampling_type=:default, kwargs...) # :bernoulli, :irrational
    for idx in first(axes(reservoir_matrix, 1)):(last(axes(reservoir_matrix, 1)) - shift)
        reservoir_matrix[idx + shift, idx] = weight[idx]
    end
    f_sample = getfield(@__MODULE__, sampling_type)
    f_sample!(rng, reservoir_matrix; kwargs...)
end

[MartinuzziFrancesco]

Not really sure why docs are failing, locally they compile just fine. @ChrisRackauckas I can't access the buildkite log it seems, is there any way to get access?

[ChrisRackauckas]

Pipeline is public now.

[MartinuzziFrancesco]

Awesome, thanks

[MartinuzziFrancesco]

I'm getting

SystemError: futimes: Disk quota exceeded

and it doesn't allow me to relaunch the workflow

[MartinuzziFrancesco]

all seems to be fine now