Calling Python Functions

yt_run_Function

int yt_run_Function( const char *function_name );

• Usage: Run Python function function_name. This is equivalent to run function_name() in Python.
• Return: YT_SUCCESS or YT_FAIL

yt_run_FunctionArguments

int yt_run_FunctionArguments( const char *function_name, int argc, ... );

• Usage: Run Python function function_name with input arguments. This API will pass total number of argc arguments. Please wrap your arguments as strings. For example, "0" for 0, "\'FieldName\'" for 'FieldName', "a" for a defined Python variable a within namespace.
• Return: YT_SUCCESS or YT_FAIL

  When using triple quotes in input arguments, use either """ or ''', but not both of them at the same time. If you really need triple quotes, stick to either one of them. For example, yt_run_FunctionArguments("func", 2, """b""", """c""") is good, but yt_run_FunctionArguments("func", 2, """b""", '''c''') is bad and leads to error.

These two API run functions inside script’s namespace, which means we can pass in objects defined in script directly.

Example

If our inline script is this:

from mpi4py import MPI

myrank = MPI.COMM_WORLD.Get_rank()
a = "var"

def func():  
    print(myrank, ": ", "Inside func()")

def funcArgs( *args ):
    print(myrank, ": ", "Inside funcArgs(", *args, ")")  # --> print funcArgs(density var 1)

We can call Python function through libyt API in simulation code:

/* libyt API: run func() in Python. */
if ( yt_run_Function( "func" ) != YT_SUCCESS ){  
    fprintf( stderr, "ERROR: func() failed!\n" );
    exit( EXIT_FAILURE );  
}

/* libyt API: run funcArgs('density', a, 1) in Python. */
if ( yt_run_FunctionArguments( "funcArgs", 3, "\'density\'", "a", "1" ) != YT_SUCCESS ){
    fprintf( stderr, "ERROR: funcArgs() failed!\n" );  
    exit( EXIT_FAILURE );  
}