Inf values for ptfrac for Zprime
I found issues with the zprime ptfrac for the current dumped p5169 samples. Some tracks have inf
values. This leads to issues in the calculation (and of course usage) of the scale/shift values. I'd guess this is due to the half precision writing but I'm not 100% sure yet. The question is how are we going to work with this?
The sample is user.alfroch.800030.e7954_s3681_r13144_p5169.tdd.EMPFlow.22_2_95.22-11-15-T131340_output.h5
. An example file with this issue is user.alfroch.31251746._000029.output.h5
Edited by Alexander Froch