Marian Ivanov

Results 64 comments of Marian Ivanov

For now, I suggest modifying the `scaleCont` and `scaleGlo` functions by using `nBins` as a parameter. A more elegant approach would be to create `sigmaCont` and `sigmaGlo` functions, and then...

Hello @f3sch, To minimize the scope of changes while maintaining the desired functionality and improving the precision of residual coding, I propose adding a new parameter: ```cpp constexpr float trackQAScaleBins{5.f};...

Thank you for your input. It seems there’s still an issue. On average, the results look fine, but the unit test is failing. The coding appears to be less precise...

## Support material for statement above: Tail and rounding error (nint) - Unit test Definition in unit test: ```c++ tree->SetAlias("qpt","trackITSTPCProp.mP[4]"); tree->SetAlias("beta","sqrt(min(50./mdEdx.dEdxMaxTPC, 1.))"); tree->SetAlias("x","qpt/beta"); for (int i=0; iSetAlias(Form("scaleCont%d",i),Form("scaleBins/sqrt(%f**2 + (%f*x)**2)",trackQAScaleContP0[i],trackQAScaleContP1[i])); tree->SetAlias(Form("scaleGlo%d",i),Form("scaleBins/sqrt(%f**2...

### Fix Scaling Issue for `dRefCont` Parameters The issue with the tails has been identified and should be fixed. The incorrect scaling factor `scaleGlo` was used instead of `scaleCont` for...

In case I used the same "buggy" value for decoding of delta everything is like expected and Unit test is OK. Normalized delta is -i nterval [-0.5,0.5] nad residuals after...

Also, other tests with the deltaGlobal P0 and deltaCont P4 are working as expected - after local fix https://github.com/AliceO2Group/AliceO2/pull/13633#issuecomment-2496163311 Test code: https://gitlab.cern.ch/alice-tpc-offline/alice-tpc-notes/-/blob/b65b22cc85a1f27cb862cd653e1c97b0cf5aabbf/JIRA/O2-5095/test_trackQADelta.C#L115-179 As an example unit test for the Glo...

### **Subject:** Delta Coding: Disk Size Impact and Full Data Size Report - Using the `tree->Print()` function, I calculated the impact of adding **2 x 5 Bytes** for delta coding...

@f3sch Great. Will you also post the updated streamer so I can perform a check? If the new check passes, I will recommend proceeding with the full request.

**Subject:** Feedback on Modifications and tpcTime0 Compression Hello @f3sch, Thank you for the modifications. The new file has passed the unit test, and the ITS+TPC deltas are now stored correctly....