I have completed the plan below. The result appears negative - adding enough
extra noise to elliminate the B2xB2-B2xRx1 discrepancy blows the Rx1xRx1 auto
spectra out of the water over the same range of angular scale as our nominal
detection.
http://bicep0.caltech.edu/~spuder/analysis_logbook/analysis/20131118_rp3
Clem
Subject: Re: [Bicep2-list] posting on B2xB1
From: John Kovac <John.Kovac.Guest(a)usap.gov>
Date: Tue, 12 Nov 2013 21:18:29 +1300
To: Clem Pryke <pryke(a)physics.umn.edu>du>, jmkovac(a)cfa.harvard.edu
Cc: Chao-Lin Kuo <clkuo(a)stanford.edu>du>, Walt Ogburn <ogburn(a)stanford.edu>du>,
bicep2-list <bicep2-list(a)lists.fas.harvard.edu>
I don't see anything since John's email on
Friday morning - which I mostly agree with.
No, no one responded. Thank you for taking it on--your plan below
sounds good.
J
On 11/12/13, 5:26 PM, Clem Pryke wrote:
I am just getting back to this - I don't see
anything since John's email on
Friday morning - which I mostly agree with.
If the hypothesis is that there is a real signal plus a systematic in Rx1
which cancels it then a concrete test plan is as follows:
- Scale the r=0.1 sims to r=0.3 and add to the signal+noise for all receivers
- In the Rx1 sig+noise sims scale the noise up until the B2xB2-B2xRx1 spectral
jack is not unlikely (i.e. the systematic is being modelled as just extra
noise).
- Look if the Rx1xRx1 auto spectrum then shows a detection - which it very well may
because it will be debiased with the *unscaled* noise which we believe to actually be
present in Rx1.
Walt/anyone: are you working on this or something similar? If not I will get on it
tomorrow morning.
Clem
--
**********************************************************************
Clem Pryke - Associate Professor - Physics
University of Minnesota,
Room 313 Tate, 116 Church Street S.E. Minneapolis, MN, 55455
Tel: 612-624-7578 Fax: 612-624-4578 email: pryke(a)physics.umn.edu
**********************************************************************