05_summarize_results_o2.sh

#!/bin/sh

# O2 https://wiki.rc.hms.harvard.edu/display/O2/Moving+from+Orchestra+to+O2


# Give data file names as arguments for the entire script

# Set slurm log directory (make sure ~ is expanded)
slurm_log_dir=`echo ./log/`

# Set number of cores to use
n_cores=1

# RAM memory per core. To avoid confusion, it's best to include a size unit, one of K, M, G, T for kilobytes, megabytes, etc. So e.g. --mem-per-cpu=8G asks for 8 GB per core.
mem_per_cpu=8G

# specified in minutes. (Format is (D-)HH:MM:SS If one colon used, runtime is specified in MM:SS.)
run_time=12:00:00

# SLURM partitions
# https://wiki.rc.hms.harvard.edu/display/O2/How+to+choose+a+partition+in+O2
partition="short"               # <12h
# partition="medium"              # <5d
# partition="priority"            # 2 job limit, 30 day limit (same as LSF priority queue)

# email notification status
mail_type="FAIL"

# email address
email="kazukiyoshida@mail.harvard.edu"


# Define a function with two arguments
RunScript() {
    # Argument 1 is data file name (invoke from the script directory)
    data_file=$1
    # Argument 2 is R script name
    r_script=$2


    # Generate slurm log file names (remove path)
    slurm_out_txt=`echo ${data_file%\.*} | sed -e "s/.*\///g"`.${r_script%\.*}.out.txt
    slurm_err_txt=`echo ${data_file%\.*} | sed -e "s/.*\///g"`.${r_script%\.*}.err.txt

    # Generate script name
    script_name=`echo ${data_file%\.*} | sed -e "s/.*\///g"`.${r_script%\.*}.slurm


    # This part creates the script
    # This flushes the file if it exists
    echo "#!/bin/bash" > ${script_name}

    echo "#SBATCH -n "${n_cores}" # Number of cores requested" >> ${script_name}
    echo "#SBATCH -N 1 # Ensure that all cores are on one machine" >> ${script_name}
    echo "#SBATCH -t "${run_time}" # Runtime in minutes" >> ${script_name}
    echo "#SBATCH -p "${partition}" # Partition to submit to" >> ${script_name}
    echo "#SBATCH --mem-per-cpu="${mem_per_cpu}" # real memory required per cpu" >> ${script_name}
    echo "#SBATCH -o "${slurm_log_dir}${slurm_out_txt}" # Standard out goes to this file" >> ${script_name}
    echo "#SBATCH -e "${slurm_log_dir}${slurm_err_txt}" # Standard err goes to this file" >> ${script_name}
    echo "#SBATCH --mail-type="${mail_type}" # Type of email notification: BEGIN, END, FAIL, ALL" >> ${script_name}
    echo "#SBATCH --mail-user="${email}" # Email to which notifications will be sent" >> ${script_name}
    echo "#SBATCH --constraint=\"scratch2\" # Require access to scratch2 file system" >> ${script_name}

    echo "# Load R" >> ${script_name}
    echo "module load gcc" >> ${script_name}
    echo "module load R" >> ${script_name}

    echo "# Configure R local package directory AFTER R has been loaded" >> ${script_name}
    echo "export R_LIBS_USER=$HOME/apps/R:$R_LIBS_USER" >> ${script_name}

    echo "# Invoke simulation runner with file name" >> ${script_name}
    echo "Rscript" ${r_script} ${data_file} ${n_cores} >> ${script_name}

    # Show script
    echo ""
    echo "Running this script"
    cat ${script_name}
    echo ""

    # This part runs the script (data file name is coded)
    echo sbatch ${script_name}
    sbatch ${script_name}

    # This part removes the script
    echo rm ${script_name}
    rm ${script_name}
}


for file in $@
do
    RunScript ${file} 05_summarize_results.R
    sleep 1
done