[b1_ana] Digitized LiD polarization at 7T data

[O. A. Rondon]

The B28 LiD digitized polarization at 7T data are here
https://twist.phys.virginia.edu/~or/b1/lid-t_7T.txt

As I said at the meeting, after I tried a log fit and thought about it
being an unbounded function, I did a new fit to the 7.55 T polarization
build up, using the bounded form
P(t>t0) = Pf*(1 - exp(c*(t - t0))) + Pi(t0)
https://twist.phys.virginia.edu/~or/b1/lidvst7t-log-exp85.pdf

I improved this fit a bit more by fixing the Pi(t0) parameter to instead
just be the data's value Pi(t0=85.433 min) = 0.1641.

The fit for the data up to 343 min (end of build up) is
Pf	0.174+/-0.0004
c	-0.00581+/-0.00002

or P(t=inf.) = Pf+Pi = 0.338

Fitting only the last ~ 50 minutes, starting at what seems to be a final
microwave frequency tweak, with Pi(t0=295.57) = 0.2869
Pf	0.405+/-0.955
c	-0.00076+/-0.00181
so the data can't constrain this later time fit.

It would be great if we could get the original data. In the mean time I
encourage all to play with the numbers in the file. One suggestion would
be to try assigning some kind of error to each P_LiD point, such as a
fixed percentage of P error, so fit quantities like chi-squared and
parameter errors are more meaningful.

Cheers,

Oscar