[Rivet] Fwd: Re: Broken e-cal strip

[David Grellscheid]

Hi Riveters!

just for information, see the emails below. Andy Pilkington tried to 
help me find the absolute position of the broken e-cal strip, which is 
relevant for ATLAS-CONF-2011-098 and, I presume, other analyses from the 
same timeframe.

As usual, it's top secret. :-/

   David


-------- Original Message --------
Subject: Re: Broken e-cal strip
Date: Wed, 03 Aug 2011 13:24:01 +0100
From: David Grellscheid <david.grellscheid at durham.ac.uk>
To: Andrew Pilkington <Andrew.Pilkington at cern.ch>

Hi Andy,

Thanks for chasing it up! Sounds like the usual ATLAS competency muddle.

> They are a bit worried about giving the information
> out because some of the earlier runs do not have the problem, and the
> problem is now fixed, so running those cuts will still not give you the
> right answer in terms of efficiency.

That's fine, but as it stands, the information in CONF-2011-098 is not
reproducible. They should at least make it explicit how many of their
events are affected, and to what extent (is it 10% loss over the full
record or the affected subset of events?). The only thing I can do in
Rivet at the moment is to throw away 10% of events randomly for the full
run.

Thanks again!

   David


> On 29 Jul 2011, at 15:35, David Grellscheid wrote:
>
>> Hi Andy,
>>
>> the conf note ID is ATLAS-CONF-2011-098, p.3 at the bottom:
>>
>>>
>>> Due to a front–end electronics failure in one of the electromagnetic
>>> calorimeter modules, a region of the calorimeter of size ∆η × ∆φ =
>>> 1.6 × 0.4 was only partially read out. Events with any jet with pT >
>>> 50 GeV in this region are rejected. The acceptance loss caused by
>>> this selection cut is about 10%.
>>>
>>
>> Thanks!
>>
>> David
>
> +++++++++++++++++++++++++
>
> Andy Pilkington
> IPPP Durham and Manchester
>
> +++++++++++++++++++++++++
>
>
>
>