We use data from three kinds triggers to estimate the relative luminosity.
1, VPDMB(tiggerID=370011) data
2, JPX( JP0 + JP1 + JP2 +AJP ) data ( PS. remove the events include (anti-)Lambda) )
3, BBCMB data

n5: event number for spinbit 5 (B-Y-)
n6: event number for spinbit 6 (B-Y+)
n9: event number for spinbit 9 (B+Y-)
n10: event number for spinbit 10 (B+Y+)

R4 = n10/n5,       R5 = n6/n5,     R6 = n9/n5

We got the results and compared them. The attachments show the plots.
relative luminosity using VPDMB events has the highest statistical precision ~0.004 , JPX statistical precision ~0.007 , BBCMB statistical precision ~0.05

We should use the relative luminosity from VPDMB or JPX events and use the difference between the two to estimate the systematic uncertainty from the relative lumonosity.