[G12] Check g12 pass1 data

[Dennis Weygand]

Every event that would have been written out does exist on a skim  
somewhere- so nothing precludes anyone from doing their own favorite  
particle id.
Skims were done just for convenience of getting enriched samples prior  
to final physics
This was done because of the tape silo crisis- previous experiments  
with large datasets, due in part to redundant skims, are writting 4  
times the input data to the silo.
g12, in contrast, has reduced this to 2x, and when you are talking 100  
TB raw data, this becomes significant.
Perhaps this should have been thought about sooner.

On Sep 1, 2009, at 6:57 AM, Marco Ripani wrote:

> Hi Dennis, everybody,
> a question and a comment/question:
> 1) Why did g12 decide to not write the unskimmed bos files ? Is it  
> because there would be an unacceptable impact on the tape silo ?
> 2) It seems me it is not good to base a skim of the data on the  
> particle ID. The particle ID can always be redone and results can  
> change. Why not to base the skim on the number of tracks only ? It  
> should produce a significant reduction in the output size already.  
> This would allow making whatever skim anybody likes based on their  
> favorite ID scheme, including making unbiased kinematic fits. This  
> is particularly relevant for the kaons. Please keep in mind that in  
> the future other groups in the collaboration may decide to analyze  
> these data again for the same or different physics and there should  
> be the minimum possible bias.
>
> Best Regards
> Marco
>
> Dennis Weygand wrote:
>> We are looking into this. The skims are based on PID, and  
>> apparently  SEB is finding kaons that PID is not- that is all we  
>> know right now.
>> It may also be that PID is finding kaons that SEB is not- I am not   
>> sure that Valery has checked for that.
>> In any event, no events are lost, but if you wish to exploit the  
>> SEB  kaons, you may have to sum over the first five skims, which is  
>> not  really a big deal, but perhaps not
>> as convenient as using a more exclusive skim. Certainly for quick   
>> checks for enhanced kaons, the skims are fine, but may nor may not  
>> be  used for final event selection.
>> Dennis
>>
>> On Aug 28, 2009, at 1:58 PM, Carlos Salgado wrote:
>>
>>
>>> Dear Valery
>>>
>>> I do not understand your statement 2.
>>> It seems to me that ALL events that have a least one kaon are in   
>>> skim 1.
>>> Isn't this true? Some of those events could be in the other skims   
>>> but if
>>> you study kaons you do not need to analyze all skims, as all your
>>> events are contained in 1!
>>>
>>> The other problem you mention is on how will we  study the   
>>> efficiency of
>>> the skim cuts if we do not have processed files w/o those cuts?  
>>> Can  we do
>>> some ramdom smaple of full data or have we already study those skim
>>> efficiencies?
>>>
>>> See you Carlos
>>>
>>> --  --  --  --  --  --  --  --  --  --  --  --  --  --  --  --   
>>> --  --
>>> Carlos W. Salgado
>>> Professor of Physics and Astronomy
>>> Director, NSU Astronomical Observatory
>>> Norfolk State University, Norfolk, VA23504        tel:   
>>> (757)823-9212/8909
>>> Jefferson Lab, Newport News, VA23606              tel: (757)  
>>> 269-7859
>>> fax: (757) 823-8817, e-mail:salgado at jlab.org or carlos at nsu.edu
>>> http://www.jlab.org/~salgado
>>> ------------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> G12 mailing list
>>> G12 at jlab.org
>>> https://mailman.jlab.org/mailman/listinfo/g12
>>>
>>
>> --
>> Dennis Weygand
>> weygand at jlab.org
>> (757) 269-5926
>>
>>
>>
>> _______________________________________________
>> G12 mailing list
>> G12 at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/g12
>>
>>
>
> -- 
>
> Marco Ripani
> Senior Scientist, Istituto Nazionale di Fisica Nucleare - Genova
> Address:
> INFN            Via Dodecaneso 33
> 16146 GENOVA - ITALY
> tel./fax 39-010-3536458
> cell. 39-328-2191138
> Jlab office room F244
> tel. 1-757-269-7266  fax 1-757-269-6273
> E-Mail: ripani at ge.infn.it
>

--
Dennis Weygand
weygand at jlab.org
(757) 269-5926