h1. 2019-03-01 amp-evolv

h2. Actions

intro (KL)

-0th paragraph add section headings-

1st paragraph needs revision; check historical details

3rd paragraph need caveat on e+e-

4th paragraph need frictional cooling

observables (CR)

-introduce cooling formula; but phrased ito amplitude (FD)- -_CR to integrate with paper (CR)_-

-remove sentence regarding 9% centile just before eq 1-

-Add a sentence regarding "human readable" what is kNN algorithm- _FD to check CR's understanding (FD)_

-Joining sentence between "observables" and "MICE"-

MICE

(CR) -contact Jason Tarrant and photo repro regarding image- _for now keep working on it amongst ourselves_

(FD) -upload schematic latex to overleaf-

(CR) remove phase space figure to MICE note; consider 2d amplitude plot instead a la Francois thesis figure 6.25

Results

(CR) -investigate the 17.5 mm bin issue esp at 10-140- _could check rebinning but low priority for now_

(CR) _francois asked by email - why is mc_reco sample size not same as reco sample size (they are supposed to be same sample_

(CR) -investigate the scraping aperture _looks like it is MC offset_-; add a vertical line corresponding to the scraping aperture; comment in text

(FD) fractional emittance is better than most significant bin as a table

(FD) -need to do the efficiency correction on fractional emittance-

(FD) -need to do the efficiency correction on density-

(CR) -Check apertures in lH2 vs None absorber; KL claims window aperture should be mostly not there in "None"-

(CR/FD) Check systematic for density in TKU 10 mm; why is SSU CC giving such large uncertainty?

h2. Notes from CM53

Try Nature before Nature physics

Fig. 1 - schematic

(CR) -Preferably plot B_{z}, or introduce an arbitrary factor of -1 after the absorber, or label as |B|-

(CR) -Note that the sigma(x) plot comes from linear beam optics calculation-

Fig. 2 - phase space 2d plots

(CR) Interesting that there is a cliff in the x-py and y-px plots. Is this diffuser aperture (or some other aperture)

(CR) If show Phase space plots, put one beam ellipse on

Fig. 3 - amplitude pdf

(CR) Add the amplitude pdf plot without corrections to the supplementary information

(CR) Plot amplitude pdf as a histogram or without horizontal errors

(CR) Afterwards, JHC pointed out that log(n) vs sqrt(A) is linear for Gaussian beam. It might just confuse people to plot it.

(CR) Vertical axis label is wrong. Normalisation needs to be explained.

Fig. 4 - pdf ratio

(CR) Make MC less visible in pdf ratio plot - it tends to dominate

(CR) Increase the MC stats (implies going to production MC, it is time)

Fig. 5 - cdf ratio

(CR) -The number of significant figures in the table is too great. Need to understand the difference between the errors for different data-

(CR) John asked for raw pdf numbers

(CR) Discussion of which bin to choose and how to quantify the level at which we observe cooling

* Soler suggested CLS method

* Others suggested looking at significance of first bin rather than most significant bin

(CR) Move the points to the bin edge, not bin centre; look-and-feel issues

(CR) Check systematic error correlations are handled correctly

(CR) General - check the systematic error is reasonable; e.g. do we need to include big error for SSU and for SSD; check we only use -ve error in SSU 

(CR) General - move to production