Dispersion calcultion and centroid

[Alea88]

Good Morning,
I am facing an unexpected difficulty in calculating residual dispersion, I hope you can help me solve it.

I have two elegant runs (let's call them run0 and run1) which are identical but for a detail: in the second run, the first RF cavity of the LINAC is changed to have a different voltage changed with &alter_elements as follows:

&alter_elements
name = RF_C_TW_0,
item = VOLT,
value = 0.9,
multiplicative = 1,
&end

I track the same beam (same initial conditions) in both runs.

I want to calculate now the difference in the centroids along the LINAC: I use the run0.cen and run1.cen files and calculate it as

CxDiff=Cx0-Cx1

Now this difference should be equal, to first order, to the dispersion times the energy difference i.e.

CxDiff=-etax*(pCentral0-pCentral1)/pCentral0

where pCentral0 is the value of the pCentral parameter read from the run0.fin file, and analogously for pCentral1.

Unfortunately when comparing the data this is not the result I get...

In attachment you find the .ele file of the run1.ele (run0.ele is identical but for the absence of &alter_elements) and the result of the postprocessing of the data done with MATLAB.
Can you help me understand where is the problem?

Many thanks,
best regards,
Giovanna

[Alea88]

Dear all,
after some deeper reading of the elegant manual I realised I had to set the first_is_fiducial=1 flag , I only now realised it was a fiducialization issue.

Best regards,
Giovanna