University of Tennessee, Knoxville TRACE: Tennessee Research and Creative Exchange Masters Theses Graduate School 8-2009 An Automated Approach to the Study and Classification of Colliding and Interacting Galaxies Steven Alvaro Boada University of Tennessee - Knoxville Follow this and additional works at: https://trace.tennessee.edu/utk_gradthes Part of the Physics Commons Recommended Citation Boada, Steven Alvaro, "An Automated Approach to the Study and Classification of Colliding and Interacting Galaxies. " Master's Thesis, University of Tennessee, 2009. https://trace.tennessee.edu/utk_gradthes/19 This Thesis is brought to you for free and open access by the Graduate School at TRACE: Tennessee Research and Creative Exchange. It has been accepted for inclusion in Masters Theses by an authorized administrator of TRACE: Tennessee Research and Creative Exchange. For more information, please contact [email protected]. To the Graduate Council: I am submitting herewith a thesis written by Steven Alvaro Boada entitled "An Automated Approach to the Study and Classification of Colliding and Interacting Galaxies." I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in partial fulfillment of the equirr ements for the degree of Master of Science, with a major in Physics. Mike Guidry, Major Professor We have read this thesis and recommend its acceptance: Jirina Stone, Robert Grzywacz Accepted for the Council: Carolyn R. Hodges Vice Provost and Dean of the Graduate School (Original signatures are on file with official studentecor r ds.) To the Graduate Council: I am submitting herewith a thesis written by Steven Alvaro Boada entitled “An Automated Approach to the Study and Classification of Colliding and Interacting Galaxies”. I have examined the final electronic copy of this thesis for form and content and recommend that it be accepted in fulfillment of the requirements for the degree of Master of Science, with a major in Physics. Mike Guidry, Major Professor We have read this thesis and recommend its acceptance: Jirina Stone Robert Grzywacz Accepted for the Council: Carolyn R. Hodges, Vice Provost and Dean of the Graduate School (Original signatures are on file with official student records.) An Automated Approach to the Study and Classification of Colliding and Interacting Galaxies A Thesis Presented for the Master of Science Degree The University of Tennessee, Knoxville Steven Alvaro Boada August 2009 Copyright c 2009 by Steven Alvaro Boada. All rights reserved. ii For JHW. iii ACKNOWLEDGMENTS The completion of this work would not have occurred without the generous support of my peers and professors. I would first like to begin by thanking The University of Tennessee Department of Physics and Astronomy. Without it, and all of the unnamed people, I would not have the mind to complete a work such as this. I would like to thank my committee members for committing their precious time to me as a student, and for their guidance over the months. My major professor, Dr. Mike Guidry, thank you for always being willing to share your guidance and wisdom with me whenever I darkened your door. Dr. Jirina Stone, thank you for your belief in me, and for all of your counsel when I most desperately needed it. Dr. Robert Grzywacz, thank you for having me a GTA all of those semesters, the good talk we shared, and for the pleasure of having you on my committee. Dr. Bronson Messer, thank you for your boundless computing assistance, and for the kind words when I became extremely frustrated. Thank you, Jay Billings, for everything. I will not attempt to list it all here. And thank you to all of my friends and family that had to listen to me talk about this every waking hour, and for loving me in spite of it. iv ABSTRACT Colliding galaxies are perhaps the greatest events changing and evolving our Universe. Consequently, the need for an understanding of how that interaction originated is very important. This thesis presents a framework in which the study of these events can be conducted in a timely and efficient manner. A genetic algorithm coupled with an initial conditions generator, a physics engine and an analysis package performs an automated search to visually match an unknown galactic interaction with a known event, thus pro- viding the starting conditions that created such an interaction. v CONTENTS 1 Prologue 1 1.1 Collision-less Collisions . ...... 3 1.2 AGeneticApproach ............................... 3 1.2.1 TheIndividualvs. Population . 4 1.2.2 MemberFitness ............................. 4 1.2.3 Evolution ................................. 4 2 Starscream 6 2.1 TheDarkMatterHalo.............................. 6 2.2 TheStarryDisks ................................. 8 2.3 TheCombination................................. 8 2.4 Initial Positions and Velocities . ...... 9 2.5 TheOrbit..................................... 10 2.6 TheN-BodyCodeGadget-2. 11 3 Genetic Algorithms 12 3.1 TheInitialPopulation . .. .. .. .. .. .. .. 12 3.2 Evaluation of the Population . 13 3.3 CrossoverandMutation . .. .. .. .. .. .. .. .. 14 3.4 Evolution of the Population . 16 3.5 MaximizingFunctions . .. .. .. .. .. .. .. .. 19 vi 4 The Workflow and Framework 21 4.1 SchematicandOverViews. 21 4.2 TheGeneticAlgorithm ............................. 23 4.2.1 The Implemented Genetic Algorithm . 23 4.2.2 OrganizationalMethods . 25 4.3 StarscreamandGadget ............................. 25 4.3.1 CreatingTheGalaxies . .. .. .. .. .. .. .. 26 4.3.2 ComputingResources . .. .. .. .. .. .. .. 27 4.4 TheAnalysis ................................... 27 4.4.1 PrototypeImages............................. 28 4.4.2 CandidateImages ............................ 29 5 Results 30 5.1 GalaxyBuilding ................................. 30 5.2 The Pearson Cross Correlation Coefficient . ...... 32 5.2.1 InitialPopulation. 34 5.2.2 Evolution ................................. 35 5.2.3 PMCCBasedFinalProduct . 36 5.3 Fast Fourier Transform based Correlation . ....... 38 5.3.1 InitialGuesses .............................. 40 5.3.2 Evolution ................................. 42 5.3.3 FFTBasedFinalProducts . 42 6 Conclusions 45 6.1 TheFramework.................................. 45 6.2 TheAnalysisPackages . .. .. .. .. .. .. .. .. 46 6.2.1 PMCC Based Evaluation . 46 6.2.2 FFT Based Evaluation . 47 6.3 FutureWork ................................... 47 Bibliography 49 vii Appendix 52 A Selected Source Code 53 A.1 populate.py .................................... 53 A.2 build ref.c..................................... 54 A.3 build candidates.c ................................ 57 A.4 compare.c ..................................... 60 Vita 62 viii LIST OF TABLES 5.1 GeneratingParameters. 31 5.2 InitialPopulationOne . .. .. .. .. .. .. .. 35 5.3 Evolved High Galaxy vs. Prototype . 37 5.4 Initial High vs. Evolved High vs. Prototype . ....... 38 5.5 InitialPopulationTwo. 41 5.6 FFT Based Evolved High vs. Prototype . 44 5.7 FFT Based Initial High vs. Evolved High vs. Prototype . ........ 44 ix LIST OF FIGURES 1.1 Sketch of the Whirl Pool Galaxy made by Lord Rosse in 1845 . ....... 2 2.1 The rotation curve for a typical spiral galaxy, where (A) is the expected rotation curve and (B) is the observed curve. .... 7 3.1 AsimpleGaussian ................................ 13 3.2 The initial population with thirty population individuals........... 14 3.3 Flow Diagram showing the evolution of a breeding population........ 17 3.4 TheSin(x)Sin(y)function. 20 4.1 Schematic View of the Workflow Including Connections . ........ 22 4.2 Schematic View of the Dataflow through the Framework . ....... 23 5.1 The prototype image, the galaxy collision shown here is the simulation that isbeingmatched.................................. 31 5.2 Low fitness (left), medium fitness (middle), high fitness (right) along with theprototypegalaxy(bottom) . 34 5.3 Evolution of the population through time. The top section is the evolution of the fitness of highest member of the breeding population, and the bot- tom section is the average fitness of that breeding population including all breedingmembers. ................................ 36 5.4 The high fitness galaxy after 100 generations (left) and the prototype galaxy (right)....................................... 37 5.5 Initial High (left), Final High (Middle), Prototype (Right).......... 38 x 5.6 Low fitness (left), medium fitness (middle), high fitness (right) along with theprototypegalaxy(bottom) . 40 5.7 Evolution of the population through time. The top section is the evolution of the fitness of the highest member of the breeding population, and the bottom section is the average fitness of that breeding population including allbreedingmembers. .............................. 43 5.8 The high fitness galaxy after 66 generations (left) and the prototype galaxy (right)....................................... 43 5.9 Initial High (left), Final High (Middle), Prototype (Right).......... 44 xi CHAPTER ONE PROLOGUE The first galaxy formed long before anyone was around to see it. In 2007, a team using the Keck Observatory discovered a set of galaxies 13.2 billion light years away, and hence forming when the universe was a mere 500 million years old [Stark et al., 2007]. The first record we have of an observed Galaxy1 is from the ancient Greek Democritus (450-370 B.C.). He was followed by another Greek, Aristotle (384-322 B.C.), several Arab astronomers, and then Galileo Galilei (1564-1642 A.D.), who in 1610 proved to the world that a Galaxy was composed of a massive number of faint stars.