This regression deals with partonic bin 15-20 GeV/c for the embedding used in the Collins paper. Performing a single partonic bin regression serves as a "proof-of-principle" to assure we are doing things correctly. The embedding code for this analysis can be found at: /star/u/rsalinas/embedding_fortmva_P.C

• Loop over detector Jets 
  • Loop over particle Jets 
    • Choose best particle jet 
  • Match to detector Jet 
  • Save particle matched and detector matched to tree 

• Currently, the events weights for each partonic bin are not implemented correctly and we don't expect a continuous falling spectrum for this individual partonic bin. The goal is to be able to get the partonic bin weights together properly to cover all partonic bins. This will be posted in the next update as another blog entry. 

• Input variables 

• Regression Target 

• From the three previous methods (check previous blog), we get the following regression:

• Deviation plots from regression and true Jet pt for the three different methods