Marissa Adams (​madams@pas.rochester.edu)

Graduated from the University of Rochester with a Physics B.S. and Mathematics B.A. From May 2014 to August 2015 I have been employed as a technical assistant in the group. In the Fall of 2015 I will be continuing my education at the University of Rochester Department of Physics & Astronomy as a PhD student (plasma physics). Here, on my page, I am documenting, as well as describing of projects I am working on. Hopefully you find it useful.

Projects

CollidingFlows

Image 1. Temporal evolution of the column density map of two colliding flows over approximately 27 Myr. Left panes are the projection of the mass down the x-axis, down the barrel of the two flows. The right panes illustrate the collision region, looking down the y-axis. Units of distance are in parsecs. Pink points are the sinks that form. The flows collide at a shear angle of 15 degrees from the normal.

With E. Kaminski, A. Frank, F. Heitsch, C. Hiag, and J. Carroll-Nellenback.

• See the ​CollidingFlows Figures for our paper.
  • My page on the particular ​E vs. t Plots featured on our figures page, mentioned above.
• Erica's ​High Resolution CollidingFlows Page.

Visualizing PN/PPN Data with SHAPE

With B. Balick, M. Huarte-Espinosa.

Under Construction

• ​Shape Wiki User Guide that I have crafted.

Turbulence Studies

With J. Carroll-Nellenback, A. Frank.

• ​WireTurbulenceStudies Page
• ​Job Statistics Page

---

Library

​My Library lists some papers that are useful for my projects. It also includes a link, and a summary of the article.

---

Scripts

Image Compressor (bash)

• Utilizes ghost script.
• Takes an image and compresses it. Good for larger vector images.
• ​Wrote a blog post on it for shits and giggles.

#!/bin/bash                                                                                                                        
#Script compressor.sh                                                                                                              

files=$@

if [ "$files" == "" ];
then
    echo "Error: No files specified, dude."
    exit
fi

for i in $files;
do
    echo \"$i\"
    gs \
        -o "${i%.pdf}-comp.pdf" \
        -sDEVICE=pdfwrite \
        -dColorConversionStrategy=/LeaveColorUnchanged \
        -dColorsampleMonoImages=false \
        -dColorsampleGrayImages=false \
        -dColorsampleColorImages=false \
        -dAutoFilterColorImages=false \
        -dAutoFilterGrayImages=false \
        -dColorImageFilter=/FlateEncode \
        -dGrayImageFilter=/FlateEncode \
        "${i}"
done

Name Changer (python)

• For post processing elements (in particular column density maps).
• Changes the name of the file easily so you don't have to do funky visualization stuff in visit.

import os
import sys

start = int(sys.argv[1])
end = int(sys.argv[2])

for i in range(start,end+1):
    inpfig="Mass_along_1_00"+str(i)+".bov"
    outfig="mass1_along_1_00"+str(i)+".bov"
    sfig = "mv %s %s" %(inpfig,outfig)
    os.system(sfig)

.okc to .csv Converter (python)

• Reads in .okc files, converts the data in the .okc to proper units, spits out a .csv so you can import the data into excel.

#!/usr/bin/env python                                                                                                              
# ^ program.py not python program.py                                                                                               

import glob
import numpy

time_scaler = 0.00825*3.81360279481384188E+14*3.16888E-8*1E-6
mass_scaler = 0.619591607894786331E+32*5.03E-34

# cd into the out directory                                                                                                        
out_dir = "./out/"

# Accessing the *.okc files and sorts them                                                                                         
lst = glob.glob(out_dir + "*.okc")
lst.sort()

with open("./s15_rawsink_data_scaled.csv", "w") as sink_f:
    sink_f.write("Time" + "," + "Core 1 Mass" + "," + "Core 2 Mass" + "," + "Core 3 Mass" + ","+ "Core 4 Mass"  + "," + "Core 5 Ma\
ss\n")

    for i in range(0,len(lst)):
        with open(lst[i]) as f:
            okc_lines = f.readlines()
            noparams, nosinks, noparams_mm = [int(x) for x in okc_lines[0].split()]
            sinkdata =[float(y[3]) for y in [x.split() for x in okc_lines[-nosinks:]]]
            sink_f.write(str(i*time_scaler) + ",")
            for j in sinkdata:
                sink_f.write(str(j*mass_scaler) + ",")
            for j in range(0,5-len(sinkdata)):
                sink_f.write("0.0" + ",")
            sink_f.write("\n")

Post Processing Submissions (bash)

• Creates post processing directories in your out directory.
• Edits and submits the jobs for your executable.
• You do have to manually move the simulation into the directory once the job has completed.

#!/bin/bash                                                                                                                        

   ## INITIALIZING INVARIANTS ##                                                                                                   

#Put the name of your executable here!                                                                                             
EXECUTABLE="astrobear_Bvn"
#Put the name of your bash script here!                                                                                            
SUBMISSION_SCRIPT="job.sh"

   ## FINDING RUN DIRECTORIES ##                                                                                                   

#/scratch/[USERNAME]/[REST OF PATH]/ Make sure script is here <--                                                                    

cd /scratch/[USERNAME]/[REST OF PATH]/

PP_PATH="Directory1, Directory2"

   ## MAKING A LOOP ##                                                                                                             

for d in ${PP_PATH}; do

    cd ${d}

      ## MAKING UNIQUE DIRECTORIES ##                                                                                              

   #Name of the directory we're going to create in...                                                                              
    DIRECTORY_NAME="${EXECUTABLE}_`echo ${d} | tr -d '/'`_`date +%s`"
   #your out directory.                                                                                                            
    OUTPUT_DIR="out/"

   #Checker to see if you have an out directory.                                                                                   
    if [ \! -e ${OUTPUT_DIR} ] || [ \! -d ${OUTPUT_DIR} ]; then
        echo "There is a problem with your ${OUTPUT_DIR} directory, please check. Exiting."
        exit
    fi

   #Makes a new directory in your out directory for post processing with the executable, run, and date in the name.                
    mkdir out/${DIRECTORY_NAME}

      ## EDITTING BATCH SCRIPT ##                                                                                                  
   #Note this change is universal for all PP_PATH                                                                                  
    sed -i 's/--reservation .*//' job.sh
    sed -i 's/-N .*/-N 32/' job.sh
    sed -i 's/--ntasks-per-node .*/--ntasks-per-node 16/' job.sh
    sed -i 's/-c .*/-c 1/' job.sh
    sed -i 's/-p .*/-p debug/' job.sh
    sed -i 's/--time .*/--time 00-00:30:00/' job.sh
    sed -i 's/#SBATCH --mail-user email1@pas.rochester.edu/##SBATCH --mail-user email1@pas.rochester.edu/' job.sh
#    sed -i 's/#SBATCH --mail-user email2@pas.rochester.edu/##SBATCH --mail-user email2@pas.rochester.edu/' job.sh                 
    sed -i "s/astrobear.* > astrobear.*.log/${EXECUTABLE} > ${EXECUTABLE}.log/" job.sh

      ## SUBMIT BATCH SCRIPT ##                                                                                                    
    sbatch ${SUBMISSION_SCRIPT}

    cd ../../
done

---

Resources

Software Documentation

ParaView:

Paraview is an visualization and analysis tool. It was developed for extremely large data sets in mind. It is open source so you can download it on your own personal machine.

• ​ParaView Homepage, where you can download the software and check for updates on their blog. They also appreciate user's requesting new features.

• ​ParaView Documentation page will immediately direct you to the ​ParaView Public Wikipage. On the wiki, you can find a ​Beginner's Tutorial.

• Download a ​presentation made by ​Martin on Paraview for the University of Houston.

SHAPE:

SHAPE is a 3D morpho-kinematical modeling application for astrophysics that has been developed by Wolfgang Steffen, Nichlas Koning, Stephan Wenger, Christophe Morisset, and Marcus Magnor. Shape is also open source and available for download. Shape allows us to use astrophysical images to inspire and develop models.

• ​Shape Webpage, where you can download, find news, publications and explore Shape's user guide.

• ​YouTube: ''Shape: 3D interactive astrophysics'', a channel dedicated to instructional videos on Shape.

• ​arXiv: ''Shape: A 3 Modeling Tool for Astrophysics'', a paper summarizing Shape's capabilities. Was accepted for publication in the IEEE Transactions on Visualization and Computer Graphics. Wolfgang Steffen, Nicholas Koning, Stephan Wenger, Christophe Morisset, Marcus Magnor.

VisIt:

VisIt, which literally stands for "Visualize It," is an open source visualization and analysis tool developed by Lawrence Livermore National Lab (LLNL). It can work on Unix, Windows and MAC workstations. It is installed on our local machines, and one can also access it in the interactive queue on BlueHive2. You can also download it yourself for free! Although you might want to check that you have a sufficient amount of ram before you visualize heavier files. Most of the visualizations we do are done with VisIt.

Lawrence Livermore National Laboratory Resources

​Download VisIt

​LLNL VisIt Manuals

AstroBEAR Documentation on particular visualization techniques one could use in VisIt.

​3D CDM Visualization Instructions : How to make a corner of a box using bov files in VisIt. Provides puts all 3 axes into one window, making data analysis easier. Also includes instructions on how to project sinks onto the construction in order to better understand where potential star formation could be occuring.

​2D Sink Curves : How to plot sinks using a .curve file instead of the P_mass object in VisIt. Much easier. Only downside is that it seems to only work for 2D data.

Supercomputer User Guides

• ​Stampede, a machine owned by the Texas Advanced Computing Center (TACC). Here is TACC's fancy ​advertisement for Stampede if you're interested.

• ​info.CIRC, a wikipage for the machines owned by Center for Integrated Research Computing (CIRC). For instance our group uses BlueHive2, Bluestreak (BS) quite heavily.

• What to do a large file transfer from one machine to another? Check out ​Globus. Here is a ​Blog Post written by Baowei on how to set up an account on Globus. Using scp and rsync may not be appropriate for data sets that are more than a few TB large.

---

Recent Blog Entries

• Location of wire simulations  -- Posted on 9 years ago : author madams
• Column Density Maps (CDM) of Wire Simulations  -- Posted on 9 years ago : author madams
• Update (07/28/2015) -- Marissa  -- Posted on 9 years ago : author madams
• some python packages that are useful for hdf5  -- Posted on 9 years ago : author madams
• Tasks going forward (07-20-2015) -- Marissa  -- Posted on 9 years ago : author madams

Page last updated at 06-02-2015.