10.2.12 Parallel_Write Procedure

The Parallel_Write Procedure takes an input character scalar or vector that is distributed across all the processors and writes it to the specified unit number. If a specific PE number is specified, only the information for that PE is written; otherwise, the information for all the PEs is written in order.

Calling syntax:

call Parallel_Write (String, Unit, PE)

Input variables:

 String  The character string (scalar or vector) to be written out, defined differently on each processor.
 PE  The processor number containing the data to be output. If not present, data from all of the processors will be output. [Optional]
 Unit  The unit number for output. If not present, unit 6 (stdout) will be used. [Optional]

The Parallel_Write code listing contains additional documentation.

Michael L. Hall