evalrcal command

evalrcal | files 
         / outfile()        random()      keepoutsiders
           time(1E-12,1E-6) dbtimesteps() verbose
           ntime(2401)
           outtimenear()    outtimeexact() outloop()
           gvalue           track
           radial rlog nr() rmax() rmin()
           spherical        intertrack    nonoyes   withage
           threads(1)       collect
           debug

Purpose

Parameters

files

For regular simulation runs:
Input listmode track file to be evaluated. If no file is specified and this command is given before the simulation is run, the listmode data produced by the physics simulation are directly passed to the chemical simulation routine.
For collecting run:
Specifies a list of gd files produced by gvalue or radial output in previous runs to be combined into a single gd file.

outfile(stem)

Stem name of the files to receive the evaluated data, in gd-format. This file is created after all listmode data have been processed.

time(tstart,tend) ntime(nt)

Do not use.

outtimenear(tlist), outtimeexact(tlist)

List of time breakpoints (in seconds) at which track and radial snapshots are to be taken.
outtimenear() selects the nearest time steps defined in dbtimesteps(), which may yield somewhat odd snapshot times.
outtimeexact() replaces the nearest time steps defined in dbtimesteps(). This may slighty alter the results, but yields exact snapshot times.

gvalue

Output G-values for the chemical species as a function of time in gd-format.
Filename:
<stem>gvalue.gd

track

Output track coordinate snap shots for the chemical species at time breakpoints specified by outtime.
Filenames e.g.:
<stem>ns001000.track.gd for the break point 1 ns.

radial

Output snap shots for the radial distance (cylindrically symmetric around z) of chemical species at time breakpoints specified by outtime.
Filenames e.g.:
<stem>ns001000.rdist.gd for the break point 1 ns.
The output is in gd-format, where the first y-column holds the raw number of species. In the second y-column this is normalized to the respective volume element. Both are normalized to the number of initial beam particles by an entry in the H: header.

spherical

For radial: calculate the spherical distance from the origin rather than the cylindrical distance from the z-axis.

nr(nr) rlog rmin(rmn) rmax(rmx)

For radial: define nr radial bins ranging from rmn to rmx, optionally with log spacing.
If not specified, some internal default will be chosen.

random(altrnd)

Selects an alternative random number generator for the chemical simulation. The alternative generator, altrnd, must be initialized using the random command. See remark below.

dbtimesteps(t1,t2,t3,...,tn, dt1,dt2,...,dtn)

Specify the time steps for the chemical simulation according to D.Boscolo's thesis.
Simulation runs from t1 through tn, with linearly spaced time intervals dt1 when in interval [t1,t2), dt2 when in interval [t2,t3) and so forth.

outloop(llist)

List of time loop counters at which track and radial snapshots are to be taken.

intertrack

Enable intertrack reactions.
By default, only reactions within the same primary particle track are considered.

nonoyes

Disable the infamous Noyes sqrt() term when the scavenging probability is calculated.
By default, it is enabled.

withage

Full Noyes term with species age. Disables nonoyes.

keepoutsiders

By default, chemical species are removed when they leave the volume in which they were created.
This option keeps them indefinitely. See remark below.

threads(nthr)

For platforms that support multithreading:
use nthr threads to process radical diffusion/reaction. This may reduce the calculation (elapsed, "wall-clock") time, see remarks below.
Multithreading is disabled for intertrack.

collect

In combination with gvalue or radial only. Different G-Value or radial distribution files, obtained by previous runs are combined into a single file specified by outfile().
This option is particularly useful to collect the output of distributed parallel runs. Note, however, that all files must strictly follow the respective output gd format and must all have exactly the same number of gd columns and data lines.
No other actions will be performed.

verbose

Some additional internal information for debugging purposes.

debug

Debug switch. Lots of output!

Remarks

• The random() option allows to run two different instances of random number generators independently, i.e. the standard one for the physics and the alternate one for the radiation chemistry simulation. This is particularly useful for development and debugging, since an altered random sequence in one simulation does not affect the other one.
• The keepoutsiders option should be used with care. Strictly speeking it is acceptable only in single volume geometries, where artificially small volumes are defined in order to keep the number of initial physics events low.
• The verbose option outputs, per time step, in this order :
  1. The current time loop number,
  2. the current time step (seconds),
  3. the # of species at the beginning of the time step (mRCal),
  4. the # of species eligible for scavenging (m1st)
  5. the # of reactions with dissolved oxygen,
  6. in version >=2410: the # of reactions with scavengers RH,
  7. in version >=2410: the # of reactions with scavengers XSH,
  8. the accumulated # of reactions with dissolved oxygen,
  9. in version >=2410: the accumulated # of reactions with scavengers RH,
  10. in version >=2410: the accumulated # of reactions with scavengers XSH,
  11. the # of species eligible for reaction (m2nd),
  12. the # of reaction distance tests (mDist2),
  13. the # of actual reactions (mReac).
• The speedup by the threads() option currently is only moderate. Even on many-core systems it rarely exceeds a factor of three. Note that the results usually are slightly different from non-threaded runs, due to the involvement of random numbers.
  Note: on AIX (at least), you need to put

   
     export MALLOCOPTIONS=multiheap
  

  in your shell startup script, otherwise multithreading will be inefficient.

Examples

More:

• Consider previous distributed parallel runs produced the output files collect.job1.gvalue.gd, collect.job2.gvalue.gd, and collect.job3.gvalue.gd, respectively,

     evalrcal collect.job?.gvalue.gd / collect gvalue outfile(collect.gvalue.gd)
  

  will combine them into a single one collect.gvalue.gd.