Once the run parameters have been set by editing the input file, the athena1.0 executable can be run using the -i option to specify the name of the input file. For example, to run the Brio & Wu shocktube use the command
% athena -i ../tst/1D-mhd/athinput.brio+wuThis will first dump out the values of all input parameters, and then when the code runs, it will print the cycle number and timestep to standard output; when it concludes it will print final diagnostic information. For example, as a simple diagnostic of code performance, the number of zones updated per cpu second is output.
A variety of command line options have been implemented in athena1.0. A list is given by the -h switch:
% athena -h Usage: ../bin/athena [options] [block/par=value ...] Options: -i <athinput_file> Specify input file [default 'athinput'] -d <run_directory> Specify run dir [default current dir] -h This Help -n Parse input, but don't run programThe -d option can be used to create a new directory in which athena1.0 will run and write the output files. The -n option is useful for debugging any parsing errors, as it will dump the contents of all parsed block/parameters.
A value for any of the valid parameter names can also be input from the command line. This, in combination with the -d option, might be useful for parameter surveys. For example, suppose a problem has three parameters, a,b,c, and athena1.0 is to be used for a survey looping over 6, 5 and 7 different values of a,b,c respectively In C-shell notation, this could be accomplished quite easily as follows:
#! /bin/csh -f foreach a (1.0 1.5 2.0 2.5 3.0 3.5) foreach b (10 20 30 40 50) foreach c (1 3 10 30 100 300 1000) set dir=run_$a_$b_$c mkdir -p $dir athena -i your_athinput -d $dir problem/a=$a problem/b=$b problem/c=$c > $dir/athena.log end end endNote that one winds up with 210 different run directories this way. An alternative would be to construct a hierarchy, with a on the top level, and b and c below. This could be accomplished with, for example, set dir=run210/$a/$b/$c.