Possible bug in MakeSoverB? about timber HOT 1 CLOSED

Did an investigation using top+H analysis. Printed bin values and then integrated manually via the following.

import math
sig = [333.0, 33.4, 23, 15.2, 10.7, 9.9]
bkg = [534.8, 78.2, 39, 54.4, 39.4, 27]

tot_s = 0
tot_b = 0
for i,s in enumerate(sig):
    b = bkg[i]
    tot_s += s
    tot_b += b
    print ('%.2f'%(tot_s/math.sqrt(tot_b)))

14.40
14.80
15.25
15.22
15.21
15.30

Of course, the hard-coded values here could be totally made up so I've attached the plot. I haven't printed the text because I couldn't get the S and B to both print without running over each other. I do promise these values match the histograms and I cross checked the printed values match those in the S/sqrt(B) pane on the bottom.
particleNet_top_cut_nminus1_Run2.pdf

from timber.