track-orbit of ID1 and ID2 OPAL-1.6 and master
In data/track-orbit, the coordinates of ID1 and ID2 are always stored. On the distribution command the user can set ID1 and ID2 by hand.
@rizzoglio_v : Please comment
Activity
I have in mind few possible situations:
- Distribution from the OPAL generator:
-
ID1 and ID2 are set by hand in the distribution command -> they will be stored in this data/track-orbit: OK
-
only ID1 is set by hand in the distribution command-> ID1 will be correctly stored in the data/track-orbit and a random particle with ID2 will be stored as well
-
ID1 and ID2 are not set by hand-> two random particles from the distribution with ID1 and ID2 are stored in the data/track-orbit -> useless output
- Distribution from a file
- Particles that in the file corresponds to ID1 and ID2 will be stored. All the three previous situations are possible
- OPAL-Cyc
- I don't know if you want to create a separate file for ID1 and ID2 or use the trackOrbit.dat (since the name you wrote was closed) that already exists. If you use the same file: are you going to replace ID0 with ID1?
In general I think it is a good solution to store ID1 and ID2 in a separate file. Above the possible scenario I have in mind are listed. This additional file should avoid to dump frequently the phase space into the H5file with a gain in computational time, size of the output and faster post-processing analysis.
-
Based on the current structure of the stat file, it is easier to store ID1 and ID2 in a separate file than in the stat file since those particles can get lost in the beamline.
Edited by Valeria Rizzoglio changed milestone to %OPAL 1.5.3
mentioned in commit ed05c51b
This is how to easily plot the trajectory of ID1 and ID2 particle using Gnuplot:
DISTRIB1: DISTRIBUTION,DISTRIBUTION = GAUSS, INPUTMOUNITS = NONE, SIGMAX = 0.0023615*fact_x, SIGMAPX = 0.00145959*gambet*fact_xd , SigmaY = 0.00612932*fact_y, SIGMAPY = 0.000837801*gambet*fact_yd , SIGMAZ = 0.00742097, SIGMAPZ = 0.00223378*gambet*fact_delp , CORRX = 0.826005, CORRY = 0.872158, CORRZ = 0.999951, R51 = -0.10973, R52 = -0.126736, R61 = -0.10805, R62 = -0.122964, **ID1={0,0,31.60060,0.0,0.0,P0}, ID2={0,0,31.60060,0.0,0.0,P0*1.01},** CUTOFFX = 5, CUTOFFPX = 5, CUTOFFY = 5, CUTOFFPY = 5, CUTOFFLONG = 5, CUTOFFPZ = 5, //move beam, starting point OFFSETX = 0, OFFSETY =0, OFFSETZ = 31.60060;test-1.in**Gnuplot command **
plot "test-1-trackOrbit.dat" using 2:(stringcolumn(1) eq "ID1"? $3:1/0) title "ID1" lc rgb "red"
Edited by adelmannmentioned in commit 0bc0cce9
changed milestone to %OPAL 1.6.0
changed milestone to %OPAL 1.9.x
-
I have compared the
trackOrbit.datfile from OPAL-T and OPAL-Cyc and there are some differences:-
Directory:
-
OPAL-T: stored in the
datadirectory -
OPAL-Cyc: stored in the simulation directory together with the other output files (stat and h5)
-
Format:
-
OPAL-T: contains an additional column SPOS
-
OPAL-Cyc: does not contain this column
-
Units:
-
OPAL-T: coordinates [m]
-
OPAL-Cyc: coordinates [mm]
-
Stored particles:
-
OPAL-T: ID1 and ID2
-
OPAL-Cyc: ID0 and ID1. But it will probably change as Andreas wrote: Because we give up on the ID=0 as reference particle with the new TracerParticles class this maybe need some adjustments.
Other difference is also mentioned in Issue #80
In general, I think, a more uniform structure for this file would be advantageous.
-
-
@rizzoglio_v: please go and write something in the manual!
-
I don't know what it is the status of this implementation with respect to the points above. In addition ID1 and ID2 seems working fine with Gaussian distribution, but I am not sure if they have been tested with the other types of distributions. I remember, Nicole had problem with her case.
In addition, I don't know the status of this implementation in OPAL-Cyc.
I would wait to write the manual till some of the points have been clarified. My entry was more a remark once the issues will be closed.
-
@gsell please comment why do you open this.
