BayesHistogram.BayesHistogram — Module

BayesHistogram.jl main procedure: function bayesianblocks( t::AbstractVector{T}; weights::AbstractVector{W} = one.(t), sumw2::AbstractVector{W} = abs2.(weights), prior = BIC(), resolution = Inf, mincounts::Real = 0, ) where {T<:Real,W<:Real}

BayesHistogram.bayesian_blocks — Method
function bayesian_blocks(
    datas::AbstractVector{T};
    weights::AbstractVector{W} = one.(t),
    sumw2::AbstractVector{W} = abs2.(weights),
    prior = BIC(),
    resolution = Inf,
    min_counts::Real = 0,
) where {T<:Real,W<:Real}
  • datas: Observations
  • weights: sample weight of each observation
  • sumw2: sum of weight^2 in each observation, this is particularly useful

when this algorihtm is used for re-binning of already made histograms where the sumw2 for each bin is different from weight^2 of each bin.

  • prior: choose from NoPrior, Pearson, Geometric, BIC, AIC, HQIC
  • resolution: handles on how fine we count along the `datas axis
  • min_counts: minimum sum of weights of a block that can be splitted.