[G12] EC Bug

[Michael C. Kunkel]

Johann,

Can you confirm this? I do not see this being addressed on the G12 
meetings page.
I find it surprising that the one time Ibrahim showed up for a meeting, 
only 3 people were there.


On 2/26/14 9:11 PM, albayrak at jlab.org wrote:
> I did mention about this and how we solved it in the G12 meeting, and I
> was not the one who solved that problem by writing a separate routine. I
> brought this up in the G12 meeting. You were not in that meeting, there
> were only three of us, Johann, myself and another person. It was Johann
> who wrote that routine and explained me how to use it. I don't know
> whether it is standard or non-standard build. I used whatever he gave me
> and followed his instructions.
>
> Regards,
>
> Ibrahim
>
>> Greetings,
>>
>> Ibrahim did not mention any bug in the EC uvw reconstruction, no did he
>> show how he worked around it with any routine.
>> The routine he uses is not in the standard build, therefore he wrote his
>> own routine to correct this and never reported it.
>>
>> If he had, I along with the other G12 lepton people, would have made it
>> a point to put it in the standard/paolone build.
>> This could cost months of simulation redoing and skimming for us.
>>
>> I respectful request to be given the member function called ECuvw() so
>> that I may put in into the build.
>>
>> BR
>> MK
>>
>> On 2/26/14 8:13 PM, Pawel Nadel-Turonski wrote:
>>> Mike,
>>>
>>> I think that Ibrahim did bring up the issue of the EC at some point. And
>>> Ibrahim’s workaround seems to work… Fixing the routine would make things
>>> a little easier, perhaps.
>>>
>>> Regarding sector 5, I think the issue is that the pedestals were not
>>> correctly set originally - and no fix was made during cooking. We had a
>>> similar issue in the second half of g13 after someone swapped out an ADC
>>> during the run. If so, there is not much one can do beyond what Ibrahim
>>> does - except to recook with better pedestals (no way of fixing it
>>> completely, though, since the lost data are lost)
>>>
>>> Cheers,
>>>
>>> 	Pawel
>>>
>>> On Feb 26, 2014, at 8:07 PM, Michael C. Kunkel <mkunkel at jlab.org> wrote:
>>>
>>>> Greetings,
>>>>
>>>> It appears that there is a bug in clasEvent. It also appears that
>>>> Ibrahim knew of this bug and never reported it.
>>>> I can say this because if I look at his code he used for the EC study,
>>>> he has a new member function that is not is the standard build of
>>>> clasEvent. The plots he shows
>>>> https://clasweb.jlab.org/rungroups/g7/wiki/images/3/36/Iha_EC_study.pdf
>>>> cannot be reproduced with the variables in the standard clasEvent build
>>>> because there is a bug in the extraction of the EC_w() (w position in
>>>> the EC).
>>>>
>>>> I have tried several workarounds, but to no avail. I can infer the dead
>>>> stips by the dead line seen in either u or v, but cannot do it for w.
>>>> As you can see from the link to Ibrahims study, sector 5 has many dead
>>>> strips in the w coordinate.
>>>>
>>>> I am attaching a plot of what I can do reconstruct in sector 5.
>>>>
>>>> Thoughts?
>>>>
>>>> BR
>>>> MK
>>>> <pim_ecuvw_phi_sec5.gif>
>