This page contains additional information and suggestions relevant to the MATLAB standalone compilation.

• The terms MCR and MATLAB are used interchangeably for SCC users because MATLAB is available and MCR shares the MATLAB path.
• As defined in the mybash.txt and mycsh.txt, the MCR path is /usr/local/matlab. This points to the system default MATLAB. If the standalone is for an MCR other than the default, load the appropriate MCR version with the module command.
• Instead of using the wrapper, you could embed the standalone-compliant data-type conversion, as shown in myStandalone.m, directly into your main program (e.g., myApp.m).
• isdeployed returns 1 if the running program is a standalone, otherwise 0 for m-file.
• For batch processing, always add an exit at the end of the main m-file to exit MATLAB and ends batch job.
• Most SCC users create standalone for batch operations. All matlab runtime options (-R) included during compilation are “burned in” to the standalone. If you use this standalone — intended for batch — in the interactive environment, it may have undesired consequences. An example is the runtime -R -nodisplay option, while makes sense for batch operations, may be undesirable in the interactive environment. Solution: create a separate standalone for interactive use, with appropriate runtime options.
• Standalone can also be made for MATLAB codes that require external C or Fortran packages. You are responsible for preparing MATLAB gateway function to serve as an interface between the calling m-file and the callee (C or Fortran code). They in turn need be compiled with the MATLAB mex compiler script and then compile with mcc with the “-a” switch to generate the standalone executable. A simple example is available.
• To submit multiple batch jobs to the SCC
  Multiple MATLAB standalone batch jobs can be submitted with qsub -t Array Job option, along with a run_standalone_job batch script . . .

  scc1$ qsub -t 100  ./run_standalone_job       <==n=SGE_TASK_ID=100
  scc1$ qsub -t 100-300:200 ./run_standalone_job <==SGE_TASK_ID=100,300

  In the above, the standalone myExecR2013a computes the sum of the arithmetic sequence    s = 1 + 2 + 3 + . . . + n. The parameter n is equated to the Array Job environment variable SGE_TASK_ID in run_standalone_job.

  run_standalone_job

  #!/bin/csh
  # SGE script for MATLAB standalone (generated with mcc)
  # Usage:
  # scc1$ qsub -t 100 ./run_standalone_job          # n=SGE_TASK_ID=100
  # scc1$ qsub -t 100-300:200 ./run_standalone_job  # SGE_TASK_ID=[100,300]
  #
  # Specify SGE batch scheduler options
  # Merge output and error files in one
  #$ -j y
  # Send email to SCC userID when job finished or aborted
  #$ -m ae
  # Request 4 cores from omp queue (example needs multicores)
  # For serial apps, prepend the line below with another #
  #$ -pe omp 4
  
  echo "\n\n********************************************"
  echo "* This job runs on $HOSTNAME"
  echo "********************************************\n\n"
  
  set NODENAME = `echo $HOSTNAME | sed 's/.scc.bu.edu//'`
  
  # Running MATLAB standalone results in creation of a cache folder
  # in home dir which may cause runtime issues. Workaround below
  echo tmpdir is $TMPDIR
  setenv MCR_CACHE_ROOT $TMPDIR
  echo "\n\n********************************************"
  echo "* Created local scratch folder /net/$NODENAME$TMPDIR"
  echo "********************************************\n\n"
  
  unsetenv DISPLAY
  
  # Output to screen go to your ?.o$JOB_ID file
  
  if ($SGE_TASK_ID == 'undefined') then
    echo "SGE_TASK_ID does not exist; use qsub -t to define it."
    exit
  else
    @ n = $SGE_TASK_ID;
  endif
  
  if ($NSLOTS == 'undefined') then
    @ nprocs = 1;         # in absence of -pe . . .
  else
    @ nprocs = $NSLOTS;   # -pe omp 4 ==> NSLOTS = 4
  endif
  
  # this standalone example expects n (parfor i=1:n) & nprocs (cores)
  ./myExecR2013a $n $nprocs