Incorrect evidence in megaplot output
The evidence (ratios) printed out in megaplot are wrong for glitchOnly runs. For example this run has
ldas-pcdev6 test_n5/trigtime_1126259462.426025391_-50610.4938965_0.0_3 $ cat evidence.dat
signal 0 0
glitch 411437.080053 215.408
noise 0 0
But the output page reports a signal-vs-glitch evidence. The reported evidence is equal to (minus) the glitch evidence, so the signal evidence being 0 is somehow ignored by megaplot.