Meeting No. 13

Vertex diagram and Adrian's comment:

I was thinking and discussing with Simon the extra vertex correction, that I have built in and I believe that it is really too much. The argument is, that probability of not emitting the soft photon is zero and contribution of this diagram to the cross-section also zero.

However, the problem of not canceling the discrepancy remains. Adrian argued, that if vertex diagram is considered, than we should also consider the leading order term in our simulation. And actually we do. Our generator throws events according to the exp(delta) correction. If only delta or exp(delta)-1 then we would have only radiative diagrams. However, here we have everything. And the same way that leading diagram is in the simulation, also the vertex diagram is in. Hence, I was double counting!

Events missing at the elastic peak:

Our main concern now ot to understand the discrepancy at the elastic line between the data and the simulation. We see too much simulation. The exactly the same effect is observed by Jan Friedrich in his thesis. Check his comments regarding the precision of the simulation! To get a feeling if this is due to the simulation or due to data I considered the Peak. Approx. simulation and investigated also spec. A data. I added also the analysis if the carbon data. For spectrometer A and Carbon data, the discrepancy can not be observed, or we see the opposite problem.

This indicates that the problem is related either to the efficiency at the center of the spectromter's acceptance or perpahs the vertex cut (optics). Since the simulation works for A it is hard to understand, why it would not work for B.

Hydrogen analysis in spec B:

Purple line is the full simulation (with the extra correction at the elastic peak). Blue is the peaking approximation. The analysis is done such, that simulation is rescaled such, that full simulation matches data. The PA simulation uses the same normalization.

1.)

2.)

3.)

Hydrogen analysis in spec B (without strict vertex cut):

4.)

Hydrogen analysis in spec A:

5.)

6.)

7.)

Carbon analysis:

8.)

9.)

10.)

Comparison of different matrices:

I have compared the delta distributions obtained with different matrices. In the analysis I used standard analysis cuts. The mind boggling thing is the disappearance of events on the edges. This is due to cut on y0. From the two figures below one can see, that cut on the middle of acceptance, causes lack of the events on the edges.

11.)

12.)

13.)

The comparison of the matrices has shown, that matrices have the ability to cause approx. 10% difference at the elastic peak. However, in my case the change was in the wrong direction.

14.)