Skip to content

Latest commit

 

History

History
42 lines (26 loc) · 4.27 KB

README.md

File metadata and controls

42 lines (26 loc) · 4.27 KB
Raw

gp_recons

This software reconstructs the wave associated with coincidences observed between antenna triggers, where a coincidence between two antennas is defined as a time interval between triggers shorter than the time it would take for an elecromagnetic wave to cover the distance between them. Two different wavefront hypothesis are considered: a plane and a spherical (ie point source) one. This program, written in C, is an adaptation of the code developped by V. Niess for the TREND reconstruction, and is based on the Fortran PORT subroutine library minimization tools.

Input files:

RRunId_coinctable.txt

Format = UnixSec; UnitId; EvtNb; CoincNb; TriggerTime

with one line per trigger, EvtNb being the trigger ID on that specific unit, and TriggerTime being given in ns wrt first unit triggered in the coincidence. This file is produced through the build_coins() function of readData.py.

positions_GP35.txt

Format: UnitID; x; y; z

x is measured Eastwards, y Northwards, z is altitude asl (to be modified, see ToDo). Note that ALL detection units have to be listed in this file and in incremental order!

The code will search for these two files in the $GP35_TXT_PATH directory, or if this variable is not defined in the system, in the gp_recons directory.

Output files:

RRunId_sphrecons.txt

Format: CoincId; UnixSec; Mult; x; y; z; t0; Chi2; Significance

One line per reconstructed coincidence. x is the Easting coordinate of the reconstructed point source, y its Northing coordinate and z its altitude (TREND conventions), ts the time of emission (wrt the first antenna trigger hence ts<0). At this stage, Chi2 and significance are not reliable. However readRecons.py provides a way to compute a Chi2 of the fit with the plot_delays() function, which is actually an implementation of the method described in TREND 2011 paper (see Fig. 6 in particular).

RRunId_planerecons.txt

Format: CoincId; UnixSec; Mult; Theta; Theta_err; Phi; Phi_err; Chi2; Significance

One line per reconstructed coincidence. Angles are given in degrees. Theta = 0° is zenith, 90° is horizon. Phi conventions still to be checked :-p. At this stage, errors, Chi2 and significance are not reliable. However readRecons.py provides a way to compute a Chi2 of the fit with the plot_delays() function, which is actually an implementation of the method described in TREND 2011 paper (see Fig. 6 in particular).

These two output files are writen in the $GP35_TXT_PATH directory, or if this variable is not defined in the system, in the gp_recons directory. The script readRecons.py provides a fast way to plot results from these two output files.

Running the program

Simply launch ./recons RunId in the gp_recons directory.

ToDo:

  1. Install gp_recons in Ulastai for a fast analysis on site. Done by Gu Junhua, March 20. Note: gp_recons does not easily compile on all plateforms (at least fails on mine).
  2. Add function in readData.py calling recons executable directly inside this script, hence making this reconstruction step transparent to the user.
  3. Presently gp_recons follows the TREND referntial convention (x <=> Easting; y = Northing; z<=>Up, azimuth counted positivily West from North. We have to modify the code to match GRAND referential conventions (ie: x <=> Northing; y <=> Westing, z <=> up; azimuth counted postively West from North; see details here). However this should be done in coordination with V. Decoene (contact on Slack: @Valentin) who also used the TREND recons code as a base for his work on GRAND angular reconstruction and thus went to a similar process already. A temporary, fast solution is to wrap the gp_recons code in a Python layer doing the conversion between GRAND and TREND conventions.