I've been trying to find alternative methods of characterizing the goodness of the fit without relying on the uncertainties quoted by root's Fit().  The trouble is that to compute any sort of shape uncertainty, I need some parameters in the fit that allow me to vary the shape, and minuit seems to choke  on the number of closely-correlated parameters even a simple assumption yields.

This was all spurred on by the desire to fulfill an inequality.  Since the BHT3coin trigger is visibly better behaved in the vernier scan than the BHT3 trigger alone, it should have less 'nongaussianity'.  Likewise, it should be more nongaussian than my test set, which is a perfect gaussian.  In terms of uncertainties, Test<BHT3coin<BHT3.  Using the original method for computng the uncertainty yielded, instead, Test<BHT3coin~BHT3, or Test<BHT3<BHT3coin, which didn't make any sense.

In the time since, I've cleaned up the code.  I don't know exactly what change caused the results to now be in line, but checking the original method now yields the expected heirarchy:

Test<1%

BHT3coin~10%

 

BHT3~25%

 

The reason I didn't notice this earlier is that I didn't expect that the code updates would make the older BHT3 fit /worse/.

 

 

results:

BHT3coin 10097097: 359/399 90%

Reference:  All bins used.

pars out(o): 399.082764 0.022029        0.020400        0.054971

Short end:  Stopping after the ith bin (peak=1)

pars out(1): 401.847717 0.021715        0.020110        0.222882

pars out(2): 388.430084 0.021905        0.015203        -2.429168

pars out(3): 161.647797 0.012209        0.010036        6.254314

pars out(4): 359.650940 0.020084        0.018441        0.533702

pars out(5): 390.459686 0.021555        0.019927        0.126473

pars out(6): 396.901154 0.021909        0.020279        0.072610

Late Start:  Starting after the ith bin (peak=1)

pars out(1): 394.876617 0.023145        0.021562        0.061657

pars out(2): 392.792755 0.023192        0.022067        0.059506

pars out(3): 312.389191 0.026607        0.025525        0.046964

pars out(4): 1914.832031        0.018260        0.017475        0.047614

pars out(5): 597.155945 0.017500        0.042118        0.035714

pars out(6): 399.082764 0.022029        0.020400        0.054971

 

 

BHT3 10097097: 369/487 76%

Reference:  All bins used.

pars out(o): 487.112427 0.022969        0.023205        1.924335

Short end:  Stopping after the ith bin (peak=1)

pars out(1): 492.252594 0.022460        0.022690        2.209939

pars out(2): 466.578064 0.020232        0.016808        -2.300184

pars out(3): 171.623108 0.011025        0.010218        9.499646

pars out(4): 369.453094 0.018183        0.018114        3.591043

pars out(5): 427.738098 0.020495        0.020565        2.653354

pars out(6): 464.155945 0.022005        0.022175        2.189004

Late Start:  Starting after the ith bin (peak=1)

pars out(1): 482.047577 0.025223        0.025758        2.075995

pars out(2): 477.436462 0.028601        0.029993        1.985783

pars out(3): 352.298065 0.045103        0.049330        1.832697

pars out(4): 254.331482 0.128873        0.135998        1.558272

pars out(5): -884800.937500     0.018841        0.015813        1.863905

pars out(6): 487.112427 0.022969        0.023205        1.924335

 

 

fake_gaussian: 1203375/1203339 ~100%

Reference:  All bins used.

pars out(o): 1203339.000000     0.750144        0.249976        1.915647

Short end:  Stopping after the ith bin (peak=1)

pars out(1): 1203322.000000     0.750166        0.249984        1.864850

pars out(2): 1205606.125000     0.770886        0.248007        29.571959

pars out(3): 1203929.375000     0.752629        0.250125        7.025809

pars out(4): 1203375.875000     0.750043        0.249999        1.753765

pars out(5): 1211479.500000     0.752811        0.250897        -9.331962

pars out(6): 1201234.750000     0.749361        0.249694        4.372316

Late Start:  Starting after the ith bin (peak=1)

pars out(1): 1200302.125000     0.748924        0.249556        5.353150

pars out(2): 1200163.000000     0.748839        0.249536        5.481222

pars out(3): 1200293.250000     0.748960        0.249554        5.395174

pars out(4): 1199098.125000     0.748406        0.249367        6.675379

pars out(5): 1203315.125000     0.750041        0.250027        1.947869

pars out(6): 1203339.000000     0.750144        0.249976        1.915647