[Hps-analysis] Runs that seems have problem with SVT flags

[Nathan Baltzell]

I’m forwarding Sebouh’s response below.

He found that svt bias values were not archived (and so not in the conditions database, yet) at the beginning of the 2015 run, and something about svt event header flag which may be expected.  Both affect the svt flags that Rafo noticed in some runs in pass7.  Rafo, what fraction of the 1.5 and 0.5 data does this affect?  If it’s a significant amount, an SVT expert may want to try to remedy this within a couple weeks when the rest of pass7 will be finished.  I remember it being said that for missing bias archive data, it was easy to tell from the noise in the data.

-Nathan


> On Dec 6, 2017, at 7:32 PM, Sebouh Paul <sebouh.paul at gmail.com> wrote:
> 
> I have found the sources of the problems with the two run ranges:
> 
> 1)  The SVT bias flag is the one that kills the first run range (5254 < Run < 5390).  All of the other flags looked correct (ie, the fraction of events rejected reflected the fraction that I would expect to fail).  In order to determine whether or not to set this flag, the EvioToLcio checks to see which time ranges in the database had good bias.  These time ranges were uploaded (I assume by Sho), using the bias values from Mya.  The bias values in Mya are listed as "undefined" for anything prior to  2015-05-05 13:44:52.875:
> 
> /cs/certified/apps/myData/1.0/bin/myData -b "2015-05-02 00:00:00" -e "2015-05-05 23:59:59" -p3 SVT:bias:top:3:v_sens -m history
> Date                    SVT:bias:top:3:v_sens
> 2015-05-02 00:00:00.000           <undefined>
> 2015-05-05 13:44:52.875               4.88863
> 2015-05-05 13:45:00.890               34.4424
> ...
> 
> What we'd need to do is figure out if the values for these runs are archived somewhere that we could access.  
> 
> 2)  The SVT event header flag was bad for the second run range (5565 < Run < 5580), while all of the other flags looked correct.  Matt Solt, could you comment on what this flag means?
> On Dec 6, 2017, at 1:06 PM, Nathan Baltzell <baltzell at jlab.org> wrote:
> 
> Looks like there are some missing runs in the database for some SVT flags.  Previously I only checked that they started at the beginning of 1.5 mm, not whether they existed for every run.  Some SVT expert will need to take a look at this.  And maybe Sebouh can take a look, he did it for 2016.  Has anyone ever really had good SVT tracks from any of these runs?
> 
> -Nathan
> 
> 
> 
>> Begin forwarded message:
>> 
>> From: Rafayel Paremuzyan <rafopar at jlab.org>
>> Subject: Runs that seems have problem with SVT flags
>> Date: December 6, 2017 at 12:10:54 PM EST
>> To: Nathan Baltzell <baltzell at jlab.org>
>> 
>> 5254 < Run < 5390
>> 5565 < Run < 5580
>> 
>> Rafo
>