Errors too large in LHCB_2016_I1454404 REF data?
There seems to be some issue with the REF data file for LHCB_2016_I1454404.
When I download it now from HEPData, the uncertainties are a factor of two smaller that the uncertainties in the REF data file bundled with Rivet 3.1.7 for some of the tables (e.g. Table 4/ d04-x01-y01
Calculating by hand, the version downloaded from HEPData seems to be correct; for example for the first bin of T4, y01, the value from HEPData is 2.09 = sqrt(2.07^2 + 0.15^2 + 0.26^2) but the value in the rivet REF file is 4.18.
Another difference in the files is that HEPData exports sys,lumi in single quotes, whereas in the REF file the name is in double quotes, which is puzzling but maybe not relevant.
Edited by Jon Butterworth