Analysis

The Postprocessing Routines

The amrex/Tools/Plotfile/ directory contains a large number of C++-based analysis routines for AMReX datasets. Many of these can be used with both MAESTROeX and the compressible astrophysics code, CASTRO.

To compile these routines, run make in the directory.

General Analysis Routines

The following routines are generally applicable for any AMReX-based plotfile. Typing the executable names without any arguments will provide usage examples.

  • fboxinfo.cpp

    Print out some basic information about the number of boxes on each refinement level and (optionally) the bounds of each of the boxes.

  • fcompare.cpp

    Compare two plotfiles, zone-by-zone to machine precision, and report the L2-norm of the error (both absolute and relative) for each variable. This assumes that the grids are identical.
    With the optional –zone_info var argument, where var is the name of a variable, it will also report the full state for the zone where var has the largest error.

  • fextract.cpp

    Extract a 1-d line through a dataset (1-, 2-, or 3-d). This works with both uniformly-gridded or AMR datasets. For multi-dimensional datasets, the coordinate direction to extract along can be specified. The line is always taken through the center of the domain. Either a single variable or all variables, along with the coordinate information, are output to a file.

  • fextrema.cpp

    Report the min and max of each variable (or only a single variable) in one or more plotfiles.

  • fsnapshot.cpp

    Create an image (PPM file) of a single variable in a plotfile. For 3-d, the slice plane through the center of the domain is specified. Separate routines exist for 2-d and 3-d datasets.

  • ftime.f90

    For each plotfile, simply print the simulation time.

  • fvarnames.f90

    Simply print out the list of variables stored in a plotfile.

Horizontal averaging

The amrex/Tools/Postprocessing/ directory contains a few more complex analysis routines for working with AMReX-based plotfiles.

Of particular use for MAESTROeX data is the HorizontalAvg.cpp routine. This will laterally average each of the variables in a plotfile (works for both 2-d and 3-d). It is written with MAESTROeX plane-parallel geometry plotfiles in mind, and the averaging is done over the coordinate direction(s) perpendicular to gravity.