Inconsistency regarding reflex statistics
There is an inconsistency of the reflex statistics comparing the table in the "merge" section of NSXTool, the generated output files & the analysis with the program Phenix:
Output files NSXTool (#unmerged/#merged): 23256 / 19062 GUI numbers NSXToll (#unmerged/#merged): 43902 / 19062 Phenix statistics. (#unmerged/#merged): 23256 / 10362
The completeness (calculated by Phenix) is therefore much lower than estimated by NSXTool. 031121_Isigma_unmerged.sca 031121_Isigma_merged.sca 031121_Isigma.txt phenix_stats_on_031121_Isigma_unmerged.txt