On Thu, 20 Jan 2005, Tancredi Botto wrote:
> Well, from genya I got a quite different answer. Do not plan to solve this
> over email
I admint I rushed a little, we go back to check the details. anyhow, I see
the change to 45 not that much for the statistics. Till now our
measurements in parallel kinematics have heavy load on T21 while the other
side is quite free of T21 "dilusion". So to some sense one might say
our T20 comes more from perpendicular kinematics. going to 45, we infact
reverse the situation, our parallel kinematics now yields purer and larger
T20 asymmetry. so maybe just play a game to see what would happen.
> > 3. a spin angle profile as seen by ed elastic. please see the attatched
> > plot. the black curve is the "scaled to 32 deg" profile. It takes the 47
> > degree profile, scaled the tangent of angle at any point by
> > tan(32)/tan(47). the purple data points are spin angle measured from ed
> > elastic by cutting the target length into intervals. the division of the
> > -20 to 20cm range is such that each interval contains roughly equal
> > counts. data seem to want to dip more. I say we should let a actual survey
> > speak.
>
> Well, if I could fit the black data I could do a weighted integral with
> the cell density
no it would be a wates of energy. I already wasted a couple Calories doing
it, so you can sit back assured this excersice only increases entropy and
CO2 in the environment. it is scaled to nominall of 32, if you integrate
to anything very different than 32, something is wrong. a game is a game,
we need a real map.
> And I also think it yields on the systematic error discussion. The
> estimation based on the error bar on the average angle may not mean much.
> It is more instructive to redo the analysis changing the spin angle
> profile in some fashion (with this map, you could systematically shift
> the field profiles +- 5 gauss up or down, which still is close to 1%
> variation and is comparable to the size one of your purple error bar).
with a real map, we'll replace the +-0.5 deg from T20 measurement. T20
will yield to be as a diagnostics as it was initially intended anyway. In
Zilu's thesis, "We estimate that the B-field direction has an accuracy
better than 1 deg. This corresponds to a relative accuracy of about 3% in
^^^^
both transversal and longitudinal fields". admittedly they operated at
different current. any insights why we can do 3 times better than NIKHEF
so we are looking forward for a 1% variation?
> thanks, we all agree I think. Now, what is the Pzz profile from your
> data ?
still working on Pzz profile.
Chi
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:32 EST