As seen in previous blog entry, the phi distribution of charged hadrons is not uniform (probably caused by HFT - one dead sector in 2016 data-taking campaign). The proposed solution was to use the generic framework (not just to solve this, but to calculate v2 using two-particle cumulants).

The histogram of non-uniform distribution was used to compute weights as (TH1D weights is jut hadron_phi histogram from previous blog entry).

weightHadron =  (weights->GetMaximum())/(weights->GetBinContent( weights->FindBin(phiHadron) ));

To check if this method works, I produced a simple weighted histogram:

hadron_phi->Fill(phiHadron);
hadron_check->Fill(phiHadron, weightHadron);

shown HERE:

Hadron phi - weighted

When the data sample is approximately the same, the distribution is flat. However, when some of the jobs have not succeeded, the distribution was not uniform (with discontinuities on places where used to be holes). This phenomena will be checked into details in the fture.