# Notes on RNNIP fc value

Which fc value should be used for the RNNIP discriminant?
A) fc = 0.018 -> same as DL1r
B) fc = 0.07 -> used in studies before DL1r fc value was determined

**if no-one has a strong preference I'd suggest to use the DL1r value**

## Comparison Plots
Reference = black: fc = 0.018
Test = blue: fc = 0.07

- dicriminant
  

- ROC curve

- Nicole also had a plot comparing the discriminant (and could make it quickly for the same range and in logscale :grinning: ) -> distributions look consistent :+1:
 
- (and in non-logscale)

- efficiency vs Lxy and vs pT plots from the Physval code are here (10k statistics): https://juhofer.web.cern.ch/juhofer/PHYSVAL/ROC_RNNIP_fc_comparisons/

## WP values (sent by Nicole, just for reference)