(i) Installation Instructions:
------------------------------

1. Download the package from http://www.aspect-sdm.org/Parallel-R/
2. Perform installation by executing the following command
	R INSTALL taskPR_ver.tar.gz
   
   If the MPI is installed in a non-default directory
   you need to provide the MPI directory path during 
   installation as shown below

   R INSTALL --configure-args=--with-mpi=/mpi/home/path 
 					taskPR_ver.tar.gz


(ii) Parallel-R Setup:
----------------------
There are two options for starting the Parallel-R engine.

1. Calling "StartPE" with the "spawn" option true.  This will
   use a MPI-2 function call to spawn the worker processes 
   and setup the Parallel-R environment.

	StartPE ()
   
   By default StartPE function has the "spawn" option set to
   true and a pair of worker processes are spawned. 

2. Without MPI-2 library support the Parallel-R setup can be
   achieved in 2 steps

   a. In the main Parallel-R process call StartPE() function
      with the spawn option "FALSE"

   b. Then startup individual R processes and execute the
      function StartWorker(). 
   
   Please refer to the StartPE() and StartWorker() documentation 
   for more information. 

      
(iii) Execution Instructions:
-----------------------------

1. Functions that needs to be run in parallel should be 
   enclosed within function PE. Note that all the parallel
   tasks should follow the following form

	a <- func(x,y,...)

	All statements must have an output!
	Assignment should only be done using "<-" or "->".
	Using "=" for assignment will NOT work.

   To execute the above function as a parallel task it should
   be input as shown below in Parallel-R

	PE(a <- func(x,y,...))

2. The function is executed by the parallel engine in the background.
   To retrieve the result, use POBJ() function call.

  For example:
    StartPE()
	PE(a <- function(x,y))
	POBJ(a)
	print(a)

  Output "a" from the above parallel task is retrieved, but NOT returned
  from the POBJ function.

