The article is out!! Here is the link:

http://dx.doi.org/10.1393/ncc/i2018-18197-1

Here is a link to the 5 min presentation:

https://agenda.infn.it/event/13852/attachments/15263/17263/favela.pdf

A couple of gifs from the presentation can be found here:

https://drive.google.com/drive/folders/1M8DYx_ynWXXOfcUI0JUo6ChWaKPU-j41?usp=sharing

Look at the bottom for poster link

Before running make sure you have python-numpy, python-matplotlib and python-termcolor installed.

$ python helium6Example.py

Runs a simple example involving 6He

The table that can be contructed (I'll automate it later) is:

p angle	p	t1	t2	sum-Q
-29.44	8.70	9.95	28.81	67.28
-38.55	3.91	13.38	30.16	67.28
23.32	10.81	28.04	8.60	67.27
36.76	2.08	30.45	14.91	67.26

energies are in MeV, Q=-19.81.

The method still needs extensive testing but at least it is conserving energy. More updates will be made but probably in a fork or another repo.

Some important clarifications:

The article explicitly refers to the Ex=18.3 in 6He and the peak associated with it in their plot is the 18.6 peak. If the program is run using the Ex=18.3 instead of the 18.9, as it was done in the previous table, then we get the following table:

p angle	p	t1	t2	sum-Q
-23.40	12.19	9.04	26.23	67.29
-38.89	1.24	18.6	28.58	67.24
19.32	13.31	25.83	8.31	67.27
26.84	0.69	27.38	19.34	67.22

An we can see, indeed that the 18 value appears as part of the t1 spectrum. Also there is a contribution to the first peak around 9 this might be the reason the first peak looks so large, there are various contributions from various channels (at least the mentioned 2).

I decided not to use the excitation in 6He of 18.3 in the article mainly for 2 reasons:

Some previous version of the program was buggy and energy was not being conserved properly. There was a 10% error and this was enough for me to interpret the 18.6 peak in the t1 spectrum differently.
I was expecting 4 solutions and making the comparison from the calculated values to the peaks values in the t1 spectrum 2 fitted nicely but the rest didn't.

Once the program was corrected (got 0.1% error), reason 2) remained. Bear in mind that I only have the published images of the used spectra, the 2 dimensional spectrums from the article seemed to fit nicely with the Ex=18.9 so I decided to use that level as the studied one.

After some discussion with the team (mainly with Dr. Cardella) it was pointed out that even though those are the "regions" where the peaks should be, we aren't saying anything about the probabilities (other than 0 if it is energetically forbidden or !=0 in case it wasn't). The height of those peaks (if any) might be under the background and therefore not clearly visible in the spectrum.

About the poster

The link is:

https://github.com/ffavela/multifrag-test/blob/master/posterIWMClearFinal.pdf

So as you can see the table is different, that's because I used E_final instead of E_final-Q. And also because the program still had the energy conservation BUG (it still has other bugs btw). So please refer to the table of the article, not the poster. Also, the caption of the spectra shown in figure 1 is obsolete so please ignore that also (BUG etc.). And last but not least the Ex=18.6 should be 18.9 (for our paper) or 18.3 (as mentioned b4, for the PRC)

In the graphical algorithm section, you may notice that the pulling can have more than one line on an eNode eventhough on all the leaf eNodes there is only one single straight line. This is because lines can be broken during the line pulling (all the way to the root eNode).

Also on the poster I call the dots points.

Any obsessive-compulsive might notice, from how the code is written, that the implementation of the algorithm is not the same as the published one (the program pulls all the lines first for example), this is because at the moment of writting the algorithm was not clear to begin with and many ideas where tried before concluding which was the most compact way of explaining it.

I highly encourage to make your own implementation. Recursion is a difficult concept, however it's more natural that we are normally aware, for example grammar has a recursive nature and we are natural intuitive grammarians from the simple fact that you are able to speak your own language (more or less) without knowing the explicit rules. That was a good reason for explaining the algorithm in terms of a simple grammar. The analogy goes a long way but I'll leave it as is.

Also about the gifs, ani.gif shows an example of a line pull for a 3 case like the 6He BRT, the masses of the constricted fragments are equal. That means that the lengths of the velocity vectors of the sibling nodes are equal. The geometric place of the fragment that splits into the two particles is as shown.

For the 4 particle case shown in animTree.gif, I just want to clarify that it was done with an energy conservation BUG, so the CM velocity vectors should not have exactly the same magnitudes.