Small discrepancy when using hadd

Hi ROOT experts.

My current analysis produces a .root file with several histograms and profile plots. I want to be able to merge several of these .root files (produced by different datasets) using hadd. The problem is I am not sure if hadd is able to handle profile plots in addition to histograms.

I tested this by running over two small datasets separately (producing plots_1.root and plots_2.root), then running over them together (producing plots_12.root), then doing:

 hadd -f plots_added.root plots_1.root plots_2.root

When I compared plots_added.root with plots_12.root, I noticed a small discrepancy between a profile plot in one versus the other.

I’m not sure if small discrepancies are to be expected, but the short of it is: does hadd work for TProfiles?

Thanks for any assistance you can offer.


If this is still an issue in v5.34/05, please fill a bug report on our JIRA page making sure to provide both the single files and the combined files.