ziotom78 / corrnoise.jl Goto Github PK

The REQUIRE file could not be found.
cc: @ziotom78

Hi,
I create the following function, according to your example:

function estimator(σ_K, slope, f_knee_Hz, τ_s, sampling_rate_Hz)
    rng = OofRNG(GaussRNG(), slope,  1/τ_s, f_knee_Hz, sampling_rate_Hz)
    data = [randoof(rng) * σ_K for i in 1:Int(τ_s * sampling_rate_Hz)+1]
end

Every time I call estimator I get the same values. This does not happen if I put the line rng = OofRNG(GaussRNG(), slope, 1/τ_s, f_knee_Hz, sampling_rate_Hz) outside the function. This happens because the seed is reinitialized every time the function is called, so the drawn sequence of pseudo-random numbers will be always the same.

This does not happen using the Distributions package, e.g.:

function estimator(mean_val, σ, τ)
    d = Normal(mean_val, σ)
    rand(d, τ)
end

where the seed can be fixed explicitly, outside the function, with Random.seed!.
So, should be possible to disentangle the seed definition from the OofRNG function? In other words, should be possible to make CorrNoise getting the seed in the same way as Distributions does?

The REQUIRE file could not be found.
cc: @ziotom78