[G12] Check g12 pass1 data

[Valery Kubarovsky]

Hi,
I made a brief check of the g12 pass1 data. The files that I analyzed
are at  /work/clas/clasg12/clasg12/pass1/bos.
I am using seb analysis code.

Two files are attached to the mail. They are identical, but with
various y scales: log and lin.
The ps files are organized as follows:

First Page.
4 pictures from first 4 skims directories.
- 1ckaon1ctr
- 2pos1neg
-3 2ctrk
-4 not_2ctr

Second page
- other
- 1lepton
- 4ctrk
- ppbar

As you know we have ONLY skim files in the output from cooking
program. Nobody did it before. Usually the analysis groups write to
the silo one main bos file with no additional cuts at all (plus some
skim files for quick access to the particular data set).

1. If you will take to the first page you will notice that the number
of charged tracks are not in a perfect agreement
with the skim scheme. For example, there are event with 0 tracks and
one track in the skim with 1 charged kaon and one charge track.
Probably it is coming from the different analysis code that I am using
(seb). The discrepancy is at the level of 2% or so.

2. More serious observation is at the Page N3 (look log version). You
will find the kaon signal in ALL 4 skim streams. However it was
supposed to appear only in the skim N1. It means actually that this
skim has no sense because you will need to analyze ALL fore skims to
pick up the events with kaons.

3. The number of K+
1 skim  -  22500 out of 104430 events
2 skim  -   1660 out of   77585 events
3 skim  -   3750 out of  185727 events
4 skim  -   2400 out of   10995 events

4. The number of K-
1 skim  -  10000 out of 104430 events
2 skim  -   2400 out of   77585 events
3 skim  -   2400 out of  185727 events
4 skim  -    600 out of   10995 events

So you see that the number of charged kaons is big in the skims N 2,3 and 4.

People who will work with another reactions (for example pi_ pi- pi+)
will need to analyze ALL skim files as well.
It is extremely inconvenient because you will need to get 5 files from
silo to analyze the data.

I propose to return back to the conventional CLAS cooking procedure
when only one bos file is written to the silo with NO skims at all.
We can produce the additional skim outputs with  more stringent cuts
to have the possibility to get quick access for the
particular reactions (for example with two kaons or 3 charge track3).

However in general there is not a big problem to make skim from the
raw cooking files. I made it many times and it did not take a huge
time.
Usually two weeks of farm time is enough to make any skim from the raw
cooking files.

Best regards,
Valery
-------------- next part --------------
A non-text attachment was scrubbed...
Name: check_pass1_log.ps
Type: application/postscript
Size: 606664 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/g12/attachments/20090828/f8983424/attachment-0002.ps 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: check_pass1_lin.ps
Type: application/postscript
Size: 592324 bytes
Desc: not available
Url : https://mailman.jlab.org/pipermail/g12/attachments/20090828/f8983424/attachment-0003.ps