Examples

Centrality Selection

centralities_selection_events partitions events by multiplicity (most central first). The bins vector defines the upper edges of the centrality classes in percent.

n1 = Oxygen()
n2 = Oxygen()

participants = Participants(n1, n2, 0.5, 2760, 1.0, 0.0)
events = rand(threaded(participants), 200_000)

bins = [10, 20, 30, 40, 50, 60, 70, 80, 90, 100]
batches = centralities_selection_events(events, bins)

central_0_10 = batches[1]
peripheral_50_60 = batches[6]

Eccentrities

epsilon_n_psi_n computes the n-th eccentricity of a given event. The following script can be used to compute e.g. the average epsilon_2 in the specified centrality classes.

m = 2
function avg_eccentricity(batch, n)
    ecc = epsilon_n_psi_n.(batch, n)
    return mean(getindex.(ecc, 1)), mean(getindex.(ecc, 2))
end
ecc2 = [avg_eccentricity(batches[i], m) for i in eachindex(batches)]

ecc2_magnitude = getindex.(ecc2, 1)
ecc2_angle = getindex.(ecc2, 2)

Background and Two-Point Correlator

This is a high-level workflow that builds the background profile and correlators. It can be expensive for large minBiasEvents.

fmGeV = 5.0
entropy(T) = 47.5 * 4 * T^3 * pi^2 / 90 * fmGeV^3
entropyToTemp = InverseFunction(entropy)
dSdT(T) = 1 / (3 * 47.5 * 4 * T^2 * pi^2 / 90 * fmGeV^2)

Norm = 100
m_list = [2,3]

bg, twpt = generate_bg_twpt_fct(entropyToTemp,dSdT,
    Norm,
    n1,n2,w ,k,p,s_NN, 
    bins, m_list; 
    minBiasEvents = 1000, r_grid = 0:1.:5, NumPhiPoints = 20, Threaded = true, Nfields = 10)

The function return an error if not enough minBiasEvents are provided for the selected amount of bins.

Background and Two-Point Correlator with File I/O

You can also generate the background profiles and two-point correlators while automatically saving them to disk using generate_bg_twpt_fct_save. This function processes each centrality bin separately, allowing it to read previously computed results and avoid recomputation.

bg, twpt = generate_bg_twpt_fct_save(entropyToTemp, dSdT,
    Norm,
    n1, n2, w ,k,p,s_NN,
    bins, m_list;
    minBiasEvents = 1000,
    r_grid = 0:1.:5,
    NumPhiPoints = 20,
    Threaded = true,
    Nfields = 10,
    path = "./results/",
    override_files = false)

The function will:

• Create a Participants object and sample events for each centrality bin.
• Write background profiles and correlators to disk using the naming convention from construct_trento_names.
• If files already exist and override_files=false, it will read from disk instead of recomputing.
• Return the full bg array and correlator tensor twpt across all bins.

This approach is useful for expensive calculations or workflows where you need persistent storage of intermediate results.