diff --git a/README.md b/README.md index 6d9c96f..27148c9 100644 --- a/README.md +++ b/README.md @@ -1,13 +1,11 @@ ![GitHub Workflow Status](https://github.com/deepskies/DeepBench/actions/workflows/test-bench.yml/badge.svg?label=test) [![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0) [![PyPI version](https://badge.fury.io/py/deepbench.svg)](https://badge.fury.io/py/deepbench) - +[![Documentation Status](https://readthedocs.org/projects/deepbench/badge/?version=latest)](https://deepbench.readthedocs.io/en/latest/?badge=latest) ### What is it? Simulation library for very simple simulations to *benchmark* machine learning algorithms. - - ### Why do we need it? Why is it useful? 1. There are very universally recognized scientifically meaningful benchmark data sets, or methods with which to generate them. 2. A very simple data set will have objects, patterns, and signals that are intuitively quanitifiable and will be fast to generate. @@ -15,9 +13,9 @@ Simulation library for very simple simulations to *benchmark* machine learning a ## Documentation -#### Readthedocs link coming soon!! +#### [ReadTheDocs](https://deepbench.readthedocs.io/en/latest/) -#### To build +#### To build from source ``` pip install sphinx cd docs diff --git a/paper/figures/00EF172E-27F9-44CD-9A48-1C16CA8882A0_1_201_a.jpeg b/paper/figures/00EF172E-27F9-44CD-9A48-1C16CA8882A0_1_201_a.jpeg deleted file mode 100644 index 82137ee..0000000 Binary files a/paper/figures/00EF172E-27F9-44CD-9A48-1C16CA8882A0_1_201_a.jpeg and /dev/null differ diff --git a/paper/figures/0F50EB6B-668C-418F-A449-5B1DD1017780_1_201_a.jpeg b/paper/figures/0F50EB6B-668C-418F-A449-5B1DD1017780_1_201_a.jpeg deleted file mode 100644 index fd629bd..0000000 Binary files a/paper/figures/0F50EB6B-668C-418F-A449-5B1DD1017780_1_201_a.jpeg and /dev/null differ diff --git a/paper/figures/2C4288C5-EF4B-4695-AA56-AA769E7A418D_1_201_a.jpeg b/paper/figures/2C4288C5-EF4B-4695-AA56-AA769E7A418D_1_201_a.jpeg deleted file mode 100644 index 318e1c5..0000000 Binary files a/paper/figures/2C4288C5-EF4B-4695-AA56-AA769E7A418D_1_201_a.jpeg and /dev/null differ diff --git a/paper/figures/2F503508-F14A-43DC-9105-3363A6AF7B45_1_201_a.jpeg b/paper/figures/2F503508-F14A-43DC-9105-3363A6AF7B45_1_201_a.jpeg deleted file mode 100644 index ef5c404..0000000 Binary files a/paper/figures/2F503508-F14A-43DC-9105-3363A6AF7B45_1_201_a.jpeg and /dev/null differ diff --git a/paper/figures/54D7B0AE-B5E3-4D15-8623-BCC00558C9DA_1_201_a.jpeg b/paper/figures/54D7B0AE-B5E3-4D15-8623-BCC00558C9DA_1_201_a.jpeg deleted file mode 100644 index 8ec2d34..0000000 Binary files a/paper/figures/54D7B0AE-B5E3-4D15-8623-BCC00558C9DA_1_201_a.jpeg and /dev/null differ diff --git a/paper/figures/80A7B764-89E1-417D-BDC3-0B4F02C139F8_1_201_a.jpeg b/paper/figures/80A7B764-89E1-417D-BDC3-0B4F02C139F8_1_201_a.jpeg deleted file mode 100644 index c7c7143..0000000 Binary files a/paper/figures/80A7B764-89E1-417D-BDC3-0B4F02C139F8_1_201_a.jpeg and /dev/null differ diff --git a/paper/figures/DeepBench.png b/paper/figures/DeepBench.png new file mode 100644 index 0000000..1854878 Binary files /dev/null and b/paper/figures/DeepBench.png differ diff --git a/paper/figures/code/example_objects.py b/paper/figures/code/example_objects.py new file mode 100644 index 0000000..21fb0ff --- /dev/null +++ b/paper/figures/code/example_objects.py @@ -0,0 +1,137 @@ +from deepbench.image import SkyImage, ShapeImage +from deepbench.physics_object import HamiltonianPendulum, Pendulum +import matplotlib.pyplot as plt +import numpy as np + + +# Each image is 480,480 +image_shape = (480, 480) + +# Total images N and figure size +fig, subplots = plt.subplots(2, 4, figsize=(12, 6)) + +# Center of all images is at 480/2, 480/2 +center = image_shape[0] / 2 + + +# Parameters for each ellipse +ellipse_params = { + "center": (center, center), + "width": 100, + "height": 200, + "fill": True, + "angle": 30, +} +shape_single = ShapeImage(image_shape, object_noise_level=0.0) +single_shape_noiseless = shape_single.combine_objects( + ["ellipse"], object_params=[ellipse_params] +) + +subplots[0, 0].imshow(single_shape_noiseless) + +# Use the same parameters to make an ellipse with noise +shape_single = ShapeImage(image_shape, object_noise_level=0.4) +shape_single_noisy = shape_single.combine_objects( + ["ellipse"], object_params=[ellipse_params] +) + +subplots[0, 1].imshow(shape_single_noisy) + +# Produce a rectangle with specified line widths +line_params = { + "center": (center + int(center / 2), center), + "width": 120, + "height": 200, + "line_width": 20, +} +shape_two = ShapeImage(image_shape, object_noise_level=0) +# Use the combine objects method to make ellipses and rectangles with the above prameters +shape_two_noiseless = shape_two.combine_objects( + ["ellipse", "rectangle"], object_params=[ellipse_params, line_params] +) + +subplots[0, 2].imshow(shape_two_noiseless) + +# Do it with a noise argument now +shape_two = ShapeImage(image_shape, object_noise_level=0.2) +shape_two_noisy = shape_single.combine_objects( + ["ellipse", "rectangle"], object_params=[ellipse_params, line_params] +) + +subplots[0, 3].imshow(shape_two_noisy) + +# Read the process with specifiations for astronomy objects +star_instance = {"radius": 100.0, "amplitude": 100.0} +star_params = {"center_x": center - int(center / 2), "center_y": center} + +galaxy_instance = {"radius": 30.0, "amplitude": 200.0, "ellipse": 0.8, "theta": 0.2} +galaxy_params = {"center_x": center, "center_y": center + int(center / 2)} +subplots[1, 0].set_ylabel("Astronomy", labelpad=8.0) + +one_image_sky = SkyImage(image_shape) +one_sky = one_image_sky.combine_objects( + ["star"], instance_params=[star_instance], object_params=[star_params] +) + +subplots[1, 0].imshow(one_sky) + + +one_sky_noise = SkyImage(image_shape, object_noise_level=0.4) +one_image_sky_noise = one_sky_noise.combine_objects( + ["star"], instance_params=[star_instance], object_params=[star_params] +) + +subplots[1, 1].imshow(one_image_sky_noise) + +one_image_sky = SkyImage(image_shape) +one_sky = one_image_sky.combine_objects( + ["star", "galaxy"], + instance_params=[star_instance, galaxy_instance], + object_params=[star_params, galaxy_params], +) + +subplots[1, 2].imshow(one_sky) + + +one_sky_noise = SkyImage(image_shape, object_noise_level=0.4) +one_image_sky_noise = one_sky_noise.combine_objects( + ["star", "galaxy"], + instance_params=[star_instance, galaxy_instance], + object_params=[star_params, galaxy_params], +) + +subplots[1, 3].imshow(one_image_sky_noise) + + +one_sky_noise = SkyImage(image_shape, object_noise_level=0.4) +one_image_sky_noise = one_sky_noise.combine_objects( + ["star", "galaxy"], + instance_params=[star_instance, galaxy_instance], + object_params=[star_params, galaxy_params], +) + +subplots[1, 3].imshow(one_image_sky_noise) + +# Y axis labels for each row +subplots[0, 0].set_ylabel("Geometry", labelpad=10.0) + +# Remove unnecessary ticks, only put them on the 100 pixel marks +# Flip the images so it starts at 0.,0. +ticks = np.linspace(0, image_shape[0], int(image_shape[0] / 100)) +for plot in subplots.ravel(): + plot.autoscale(tight=True) + plot.set_yticks(ticks.tolist()[::-1]) + plot.invert_yaxis() + plot.set_xticks(ticks) + +# All object titles +subplots[0, 0].set_title("Noiseless Single Object") +subplots[0, 2].set_title("Noiseless Multi-Object") +subplots[0, 1].set_title("Noisy Single Object") +subplots[0, 3].set_title("Noisy Multi-Object") + +# Scale information +fig.supxlabel("pixel") +fig.supylabel("pixel") + +plt.savefig("../example_objects.png") diff --git a/paper/figures/code/example_pendulums.py b/paper/figures/code/example_pendulums.py new file mode 100644 index 0000000..a58d0f3 --- /dev/null +++ b/paper/figures/code/example_pendulums.py @@ -0,0 +1,89 @@ +from deepbench.physics_object import HamiltonianPendulum, Pendulum +import matplotlib.pyplot as plt +import numpy as np + + +# Define the number of objects in the plot and the total figure size +fig, subplots = plt.subplots(1, 2, figsize=(int(19 * (3 / 4)), int(7 * 3 / 4))) + +# Set the times to calculate the pendulum position over +# 1 point every second, for 0 to 25 seconds +time = np.array(np.linspace(0, 25, 25)) + +# Produce pendulum object +pendulum = Pendulum( + pendulum_arm_length=10.0, + starting_angle_radians=np.pi / 4, + acceleration_due_to_gravity=9.8, + noise_std_percent={ + "pendulum_arm_length": 0.0, + "starting_angle_radians": 0.1, + "acceleration_due_to_gravity": 0.1, + }, +) + +# Use the noiseless argument to make the pendulum w/o noise +# Plot that against the time and with scatter and line options +pendulum_noiseless = pendulum.create_object(time, noiseless=True) +subplots[0].plot(time, pendulum_noiseless, color="black") +subplots[0].scatter(time, pendulum_noiseless, color="black", label="Noiseless") + +# Use the noiseless=False to do the same with a noiseless pendulum +pendulum_noisy = pendulum.create_object(time, noiseless=False) +subplots[0].plot(time, pendulum_noisy, color="red") +subplots[0].scatter(time, pendulum_noisy, color="red", label="Noisy") + + +# Produce noiseless pendulum object for the H +pendulum = HamiltonianPendulum( + pendulum_arm_length=10.0, + starting_angle_radians=np.pi / 4, + acceleration_due_to_gravity=9.8, + noise_std_percent={ + "pendulum_arm_length": 0.0, + "starting_angle_radians": 0.0, + "acceleration_due_to_gravity": 0.0, + }, +) + +# Cacluate the pendulum positions and engeries +pendulum_data = pendulum.create_object(time) + +# Plot the line and scatterplot versions of the position wrt time +subplots[1].plot(pendulum_data[4], pendulum_data[0], color="black") +subplots[1].scatter( + pendulum_data[4], pendulum_data[0], color="black", label="Noiseless" +) + +# Repeat the process with the noisely pendulum +pendulum = HamiltonianPendulum( + pendulum_arm_length=10.0, + starting_angle_radians=np.pi / 4, + acceleration_due_to_gravity=9.8, + noise_std_percent={ + "pendulum_arm_length": 0.2, + "starting_angle_radians": 0.0, + "acceleration_due_to_gravity": 0.0, + }, +) + +pendulum_data = pendulum.create_object(time) + +subplots[1].plot(pendulum_data[4], pendulum_data[0], color="red") +subplots[1].scatter(pendulum_data[4], pendulum_data[0], color="red", label="Noisy") + +# Set plot labels +subplots[0].set_title("Newtonian") +subplots[1].set_title("Hamiltonian") + +# Set axices labels +for plot in subplots.ravel(): + # plot.set(xticks=[], yticks=[]) + + plot.set_xlabel("Time (s)") + plot.set_ylabel("X Position") + +# Assign legend location +subplots[1].legend(loc="center left", bbox_to_anchor=(1.02, 1)) + +plt.savefig("../pendulums.png") diff --git a/paper/figures/example_objects.png b/paper/figures/example_objects.png new file mode 100644 index 0000000..93055d5 Binary files /dev/null and b/paper/figures/example_objects.png differ diff --git a/paper/figures/overview_diagram.png b/paper/figures/overview_diagram.png deleted file mode 100644 index 80b014b..0000000 Binary files a/paper/figures/overview_diagram.png and /dev/null differ diff --git a/paper/figures/pendulums.png b/paper/figures/pendulums.png new file mode 100644 index 0000000..c8cfabf Binary files /dev/null and b/paper/figures/pendulums.png differ diff --git a/paper/paper.bib b/paper/paper.bib index d186879..d168256 100644 --- a/paper/paper.bib +++ b/paper/paper.bib @@ -1,194 +1,659 @@ -@Article{deeplenstronomy, - doi = {10.21105/joss.02854}, - url = {https://doi.org/10.21105/joss.02854}, +@article{abbottDARKENERGYSURVEY, + title = {{{THE DARK ENERGY SURVEY DATA RELEASE}}}, + author = {Abbott, T M C and Abdalla, F B and Allam, S and Amara, A and Annis, J and Asorey, J and Avila, S and Ballester, O and Banerji, M and Barkhouse, W and Baruah, L and Baumer, M and Bechtol, K and Becker, M R and {Benoit-L{\'e}vy}, A and Bernstein, G M and Bertin, E and Blazek, J and Bocquet, S and Brooks, D and Brout, D and {Buckley-Geer}, E and Burke, D L and Busti, V and Campisano, R and {Cardiel-Sas}, L and Rosell, A Carnero and Kind, M Carrasco and Carretero, J and Castander, F J and Cawthon, R and Chang, C and Conselice, C and Costa, G and Crocce, M and Cunha, C E and D'Andrea, C B and Costa, L N Da and Das, R and Daues, G and Davis, T M and Davis, C and Vicente, J De and Depoy, D L and Derose, J and Desai, S and Diehl, H T and Dietrich, J P and Dodelson, S and Doel, P and {Drlica-Wagner}, A and Eifler, T F and Elliott, A E and Evrard, A E and Farahi, A and Neto, A Fausti and Fernandez, E and Finley, D A and Flaugher, B and Foley, R J and Fosalba, P and Friedel, D N and Frieman, J and {Garc{\'i}a-Bellido}, J and Gaztanaga, E and Gerdes, D W and Giannantonio, T and Gill, M S S and Glazebrook, K and Goldstein, D A and Gower, M and Gruen, D and Gruendl, R A and Gschwend, J and Gupta, R R and Gutierrez, G and Hamilton, S and Hartley, W G and Hinton, S R and Hislop, J M and Hollowood, D and Honscheid, K and Hoyle, B and Huterer, D and Jain, B and James, D J and Jeltema, T and Johnson, M W G and Johnson, M D and Kacprzak, T and Kent, S and Khullar, G and Klein, M and Kovacs, A and Koziol, A M G and Krause, E and Kremin, A and Kron, R and Kuehn, K and Kuhlmann, S and Kuropatkin, N and Lahav, O and Lasker, J and Li, T S and Li, R T and Liddle, A R and Lima, M and Lin, H and {L{\'o}pez-Reyes}, P and Maccrann, N and Maia, M A G and Maloney, J D and Manera, M and March, M and Marriner, J and Marshall, J L and Martini, P and Mcclintock, T and Mckay, T and Mcmahon, R G and Melchior, P and Menanteau, F and Miller, C J and Miquel, R and Mohr, J J and Morganson, E and Mould, J and Neilsen, E and Nichol, R C and Nogueira, F and Nord, B and Nugent, P and Nunes, L and Ogando, R L C and Old, L and Pace, A B and Palmese, A and {Paz-Chinch{\'o}n}, F and Peiris, H V and Percival, W J and Petravick, D and Plazas, A A and Poh, J and Pond, C and Porredon, A and Pujol, A and Refregier, A and Reil, K and Ricker, P M and Rollins, R P and Romer, A K and Roodman, A and Rooney, P and Ross, A J and Rykoff, E S and Sako, M and Sanchez, M L and Sanchez, E and Santiago, B and Saro, A and Scarpine, V and Scolnic, D and Serrano, S and {Sevilla-Noarbe}, I and Sheldon, E and Shipp, N and Silveira, M L and Smith, M and Smith, R C and Smith, J A and {Soares-Santos}, M and Sobreira, F and Song, J and Stebbins, A and Suchyta, E and Sullivan, M and Swanson, M E C and Tarle, G and Thaler, J and Thomas, D and Thomas, R C and Troxel, M A and Tucker, D L and Vikram, V and Vivas, A K and Walker, A R and Wechsler, R H and Weller, J and Wester, W and Wolf, R C and Wu, H and Yanny, B and Zenteno, A and Zhang, Y and Zuntz, J and Juneau, S and Fitzpatrick, M and Nikutta, R and Nidever, D and Olsen, K and Scott, A}, + langid = {english}, + keywords = {/unread}, + file = {/Users/nord/Zotero/storage/3QQ7R7SJ/Abbott et al. - THE DARK ENERGY SURVEY DATA RELEASE.pdf} +} + +@misc{andreopoulosGENIENeutrinoMonte2015, + title = {The {{GENIE Neutrino Monte Carlo Generator}}: {{Physics}} and {{User Manual}}}, + shorttitle = {The {{GENIE Neutrino Monte Carlo Generator}}}, + author = {Andreopoulos, Costas and Barry, Christopher and Dytman, Steve and Gallagher, Hugh and Golan, Tomasz and Hatcher, Robert and Perdue, Gabriel and Yarba, Julia}, + year = {2015}, + month = oct, + number = {arXiv:1510.05494}, + eprint = {1510.05494}, + primaryclass = {hep-ex, physics:hep-ph}, + publisher = {{arXiv}}, + doi = {10.48550/arXiv.1510.05494}, + urldate = {2023-09-25}, + abstract = {GENIE is a suite of products for the experimental neutrino physics community. This suite includes i) a modern software framework for implementing neutrino event generators, a state-of-the-art comprehensive physics model and tools to support neutrino interaction simulation for realistic experimental setups (the Generator product), ii) extensive archives of neutrino, charged-lepton and hadron scattering data and software to produce a comprehensive set of data/MC comparisons (the Comparisons product), and iii) a generator tuning framework and fitting applications (the Tuning product). This book provides the definite guide for the GENIE Generator: It presents the software architecture and a detailed description of its physics model and official tunes. In addition, it provides a rich set of data/MC comparisons that characterise the physics performance of GENIE. Detailed step-by-step instructions on how to install and configure the Generator, run its applications and analyze its outputs are also included.}, + archiveprefix = {arxiv}, + keywords = {/unread,High Energy Physics - Experiment,High Energy Physics - Phenomenology}, + file = {/Users/nord/Zotero/storage/KIXG5XUP/Andreopoulos et al. - 2015 - The GENIE Neutrino Monte Carlo Generator Physics .pdf;/Users/nord/Zotero/storage/NR4W2BU5/1510.html} +} + +@article{astropycollaborationAstropyCommunityPython2013a, + title = {Astropy: {{A}} Community {{Python}} Package for Astronomy}, + author = {{Astropy Collaboration} and Robitaille, T. P. and Tollerud, E. J. and Greenfield, P. and Droettboom, M. and Bray, E. and Aldcroft, T. and Davis, M. and Ginsburg, A. and {Price-Whelan}, A. M. and Kerzendorf, W. E. and Conley, A. and Crighton, N. and Barbary, K. and Muna, D. and Ferguson, H. and Grollier, F. and Parikh, M. M. and Nair, P. H. and Unther, H. M. and Deil, C. and Woillez, J. and Conseil, S. and Kramer, R. and Turner, J. E. H. and Singer, L. and Fox, R. and Weaver, B. A. and Zabalza, V. and Edwards, Z. I. and Azalee Bostroem, K. and Burke, D. J. and Casey, A. R. and Crawford, S. M. and Dencheva, N. and Ely, J. and Jenness, T. and Labrie, K. and Lim, P. L. and Pierfederici, F. and Pontzen, A. and Ptak, A. and Refsdal, B. and Servillat, M. and Streicher, O.}, + year = {2013}, + month = oct, + journal = {\aa p}, + volume = {558}, + pages = {A33}, + doi = {10.1051/0004-6361/201322068}, + keywords = {/unread,methods: data analysis,methods: miscellaneous,virtual observatory tools} +} + +@article{astropycollaborationAstropyProjectBuilding2018a, + title = {The {{Astropy Project}}: {{Building}} an {{Open-science Project}} and {{Status}} of the v2.0 {{Core Package}}}, + author = {{Astropy Collaboration} and {Price-Whelan}, A. M. and Sip{\textbackslash}Hocz, B. M. and G{\"u}nther, H. M. and Lim, P. L. and Crawford, S. M. and Conseil, S. and Shupe, D. L. and Craig, M. W. and Dencheva, N. and Ginsburg, A. and {Vand erPlas}, J. T. and Bradley, L. D. and {P{\'e}rez-Su{\'a}rez}, D. and {de Val-Borro}, M. and Aldcroft, T. L. and Cruz, K. L. and Robitaille, T. P. and Tollerud, E. J. and Ardelean, C. and Babej, T. and Bach, Y. P. and Bachetti, M. and Bakanov, A. V. and Bamford, S. P. and Barentsen, G. and Barmby, P. and Baumbach, A. and Berry, K. L. and Biscani, F. and Boquien, M. and Bostroem, K. A. and Bouma, L. G. and Brammer, G. B. and Bray, E. M. and Breytenbach, H. and Buddelmeijer, H. and Burke, D. J. and Calderone, G. and Cano Rodr{\'i}guez, J. L. and Cara, M. and Cardoso, J. V. M. and Cheedella, S. and Copin, Y. and Corrales, L. and Crichton, D. and D'Avella, D. and Deil, C. and Depagne, {\'E}. and Dietrich, J. P. and Donath, A. and Droettboom, M. and Earl, N. and Erben, T. and Fabbro, S. and Ferreira, L. A. and Finethy, T. and Fox, R. T. and Garrison, L. H. and Gibbons, S. L. J. and Goldstein, D. A. and Gommers, R. and Greco, J. P. and Greenfield, P. and Groener, A. M. and Grollier, F. and Hagen, A. and Hirst, P. and Homeier, D. and Horton, A. J. and Hosseinzadeh, G. and Hu, L. and Hunkeler, J. S. and Ivezi{\'c}, {\v Z}. and Jain, A. and Jenness, T. and Kanarek, G. and Kendrew, S. and Kern, N. S. and Kerzendorf, W. E. and Khvalko, A. and King, J. and Kirkby, D. and Kulkarni, A. M. and Kumar, A. and Lee, A. and Lenz, D. and Littlefair, S. P. and Ma, Z. and Macleod, D. M. and Mastropietro, M. and McCully, C. and Montagnac, S. and Morris, B. M. and Mueller, M. and Mumford, S. J. and Muna, D. and Murphy, N. A. and Nelson, S. and Nguyen, G. H. and Ninan, J. P. and N{\"o}the, M. and Ogaz, S. and Oh, S. and Parejko, J. K. and Parley, N. and Pascual, S. and Patil, R. and Patil, A. A. and Plunkett, A. L. and Prochaska, J. X. and Rastogi, T. and Reddy Janga, V. and Sabater, J. and Sakurikar, P. and Seifert, M. and Sherbert, L. E. and {Sherwood-Taylor}, H. and Shih, A. Y. and Sick, J. and Silbiger, M. T. and Singanamalla, S. and Singer, L. P. and Sladen, P. H. and Sooley, K. A. and Sornarajah, S. and Streicher, O. and Teuben, P. and Thomas, S. W. and Tremblay, G. R. and Turner, J. E. H. and Terr{\'o}n, V. and {van Kerkwijk}, M. H. and {de la Vega}, A. and Watkins, L. L. and Weaver, B. A. and Whitmore, J. B. and Woillez, J. and Zabalza, V. and {Astropy Contributors}}, + year = {2018}, + month = sep, + journal = {\textbackslash aj}, + volume = {156}, + number = {3}, + pages = {123}, + doi = {10.3847/1538-3881/aabc4f}, + keywords = {/unread,Astrophysics - Instrumentation and Methods for Astrophysics,methods: data analysis,methods: miscellaneous,methods: statistical,reference systems} +} + +@article{birrerLenstronomyMultipurposeGravitational2018a, + title = {Lenstronomy: {{Multi-purpose}} Gravitational Lens Modelling Software Package}, + author = {Birrer, Simon and Amara, Adam}, + year = {2018}, + month = dec, + journal = {Physics of the Dark Universe}, + volume = {22}, + pages = {189--201}, + publisher = {{Elsevier BV}}, + doi = {10.1016/j.dark.2018.11.002}, + keywords = {/unread} +} + +@article{bridleHandbookGREAT08Challenge2009a, + title = {Handbook for the {{GREAT08 Challenge}}: {{An}} Image Analysis Competition for Cosmological Lensing}, + author = {Bridle, Sarah and Gill, Mandeep and Heavens, Alan and Heymans, Catherine and High, F. William and Hoekstra, Henk and Jarvis, Mike and Kirk, Donnacha and Kitching, Thomas and Kneib, Jean-Paul and Kuijken, Konrad and {Shawe-Taylor}, John and Lagatutta, David and Mandelbaum, Rachel and Massey, Richard and Mellier, Yannick and Moghaddam, Baback and Moudden, Yassir and Nakajima, Reiko and {Paulin-Henriksson}, Stephane and Pires, Sandrine and Rassat, Anais and Amara, Adam and Refregier, Alexandre and Rhodes, Jason and Schrabback, Tim and Semboloni, Elisabetta and Shmakova, Marina and van Waerbeke, Ludovic and Witherick, Dugan and Voigt, Lisa and Wittman, David and Applegate, Douglas and Balan, Sreekumar T. and Berge, Joel and Bernstein, Gary and Dahle, Hakon and Erben, Thomas}, + year = {2009}, + month = mar, + journal = {The Annals of Applied Statistics}, + volume = {3}, + number = {1}, + publisher = {{Institute of Mathematical Statistics}}, + doi = {10.1214/08-aoas222}, + keywords = {/unread} +} + +@misc{CAMBInfo, + title = {{{CAMB}}.Info}, + urldate = {2023-09-25}, + howpublished = {https://camb.info/}, + keywords = {/unread}, + file = {/Users/nord/Zotero/storage/AX5WTUUH/camb.info.html} +} + +@article{chacon-cardonaMillenniumSimulationDark2012, + title = {Millennium {{Simulation Dark Matter Haloes}}: {{Multi-fractal}} and {{Lacunarity Analysis}} with {{Homogeneity Transition}}}, + shorttitle = {Millennium {{Simulation Dark Matter Haloes}}}, + author = {{Chac{\'o}n-Cardona}, C{\'e}sar A. and {Casas-Miranda}, Rigoberto A.}, + year = {2012}, + month = dec, + journal = {Monthly Notices of the Royal Astronomical Society}, + volume = {427}, + number = {3}, + eprint = {1209.2637}, + primaryclass = {astro-ph, physics:gr-qc}, + pages = {2613--2624}, + issn = {00358711, 13652966}, + doi = {10.1111/j.1365-2966.2012.22095.x}, + urldate = {2023-08-14}, + abstract = {We investigate from the fractal viewpoint the way in which the dark matter is grouped at z = 0 in the Millennium dark matter cosmological simulation. The determination of the cross to homogeneity in the Millennium Simulation data is described from the behaviour of the fractal dimension and the lacunarity. We use the sliding window technique to calculate the fractal mass-radius dimension, the pre-factor F and the lacunarity of this fractal relation. Besides, we determinate the multi-fractal dimension and the lacunarity spectrum, including their dependence with radial distance. This calculations show a radial distance dependency of all the fractal quantities, with heterogeneity clustering of dark matter haloes up to depths of 100 Mpc/h. The dark matter haloes clustering in the Millennium Simulation shows a radial distance dependency, with two regions clearly defined. The lacunarity spectrum for values of the structure parameter q {$>$}= 1 shows regions with relative maxima, revealing the formation of clusters and voids in the dark matter haloes distribution. With the use of the multi-fractal dimension and the lacunarity spectrum, the transition to homogeneity at depths between 100 Mpc/h and 120 Mpc/h for the Millennium Simulation dark matter haloes is detected.}, + archiveprefix = {arxiv}, + keywords = {/unread,Astrophysics - Cosmology and Nongalactic Astrophysics,General Relativity and Quantum Cosmology}, + file = {/Users/nord/Zotero/storage/B6DWM2SJ/Chacón-Cardona and Casas-Miranda - 2012 - Millennium Simulation Dark Matter Haloes Multi-fr.pdf;/Users/nord/Zotero/storage/ZA8PQICN/1209.html} +} + +@article{dengMnistDatabaseHandwritten2012c, + title = {The Mnist Database of Handwritten Digit Images for Machine Learning Research}, + author = {Deng, Li}, + year = {2012}, + journal = {IEEE Signal Processing Magazine}, + volume = {29}, + number = {6}, + pages = {141--142}, + publisher = {{IEEE}}, + doi = {10.1109/MSP.2012.2211477}, + keywords = {/unread} +} + +@article{flaugherDarkEnergySurvey2005a, + title = {The {{Dark Energy Survey}}}, + author = {Flaugher, Brenna}, + year = {2005}, + month = jan, + journal = {International Journal of Modern Physics A}, + volume = {20}, + pages = {3121--3123}, + issn = {0217-751X}, + doi = {10.1142/S0217751X05025917}, + urldate = {2023-09-25}, + abstract = {Dark Energy is the dominant constituent of the universe and we have little understanding of it. We describe a new project aimed at measuring the dark energy equation of state parameter, w, to a statistical precision of \textasciitilde 5\% with four separate techniques. The survey will image 5000 deg2 in the southern sky and collect 300 million galaxies, 30,000 galaxy clusters, and 2000 Type Ia supernovae. The survey will be carried out using a new 3 deg2 mosaic camera mounted at the prime focus of the 4m Blanco telescope at CTIO.}, + keywords = {/unread,Dark energy,galaxies,supernovae}, + annotation = {ADS Bibcode: 2005IJMPA..20.3121F} +} + +@article{fluxaPixelSpaceConvolution2020, + title = {Pixel Space Convolution for Cosmic Microwave Background Experiments}, + author = {Flux{\'a}, P. and Brewer, M. K. and D{\"u}nner, R.}, + year = {2020}, + month = feb, + journal = {J. Cosmol. Astropart. Phys.}, + volume = {2020}, + number = {02}, + eprint = {1908.05662}, + primaryclass = {astro-ph}, + pages = {030--030}, + issn = {1475-7516}, + doi = {10.1088/1475-7516/2020/02/030}, + urldate = {2023-09-25}, + abstract = {Cosmic microwave background experiments have experienced an exponential increase in complexity, data size and sensitivity. One of the goals of current and future experiments is to characterize the B-mode power spectrum, which would be considered a strong evidence supporting inflation. The signal associated with inflationary B-modes is very weak, and so a successful detection requires exquisite control over systematic effects, several of which might arise due to the interaction between the electromagnetic properties of the telescope beam, the scanning strategy and the sky model. In this work, we present the Pixel Space COnvolver (PISCO), a new software tool capable of producing mock data streams for a general CMB experiment. PISCO uses a fully polarized representation of the electromagnetic properties of the telescope. PISCO also exploits the massively parallel architecture of Graphic Processing Units to accelerate the main calculation. This work shows the results of applying PISCO in several scenarios, included a realistic simulation of an ongoing experiment, the Cosmology Large Angular Scale Surveyor.}, + archiveprefix = {arxiv}, + keywords = {/unread,Astrophysics - Instrumentation and Methods for Astrophysics}, + file = {/Users/nord/Zotero/storage/F38ZZ4MP/Fluxá et al. - 2020 - Pixel space convolution for cosmic microwave backg.pdf;/Users/nord/Zotero/storage/5YLASFJL/1908.html} +} + +@misc{gonzalezGalaxyDetectionIdentification2018a, + title = {Galaxy Detection and Identification Using Deep Learning and Data Augmentation}, + author = {Gonz{\'a}lez, Roberto E. and Mu{\~n}oz, Roberto P. and Hern{\'a}ndez, Cristian A.}, + year = {2018}, + publisher = {{arXiv}}, + doi = {10.48550/ARXIV.1809.01691}, + copyright = {arXiv.org perpetual, non-exclusive license}, + keywords = {/unread,FOS: Physical sciences,Instrumentation and Methods for Astrophysics (astro-ph.IM)} +} + +@article{harrisArrayProgrammingNumPy2020b, + title = {Array Programming with {{NumPy}}}, + author = {Harris, Charles R. and Millman, K. Jarrod and van der Walt, St{\'e}fan J. and Gommers, Ralf and Virtanen, Pauli and Cournapeau, David and Wieser, Eric and Taylor, Julian and Berg, Sebastian and Smith, Nathaniel J. and Kern, Robert and Picus, Matti and Hoyer, Stephan and van Kerkwijk, Marten H. and Brett, Matthew and Haldane, Allan and del R{\'i}o, Jaime Fern{\'a}ndez and Wiebe, Mark and Peterson, Pearu and {G{\'e}rard-Marchant}, Pierre and Sheppard, Kevin and Reddy, Tyler and Weckesser, Warren and Abbasi, Hameer and Gohlke, Christoph and Oliphant, Travis E.}, + year = {2020}, + month = sep, + journal = {Nature}, + volume = {585}, + number = {7825}, + pages = {357--362}, + publisher = {{Springer Science and Business Media LLC}}, + doi = {10.1038/s41586-020-2649-2}, + keywords = {/unread} +} + +@article{heDeepResidualLearning2015d, + title = {Deep {{Residual Learning}} for {{Image Recognition}}}, + author = {He, Kaiming and Zhang, Xiangyu and Ren, Shaoqing and Sun, Jian}, + year = {2015}, + journal = {CoRR}, + volume = {abs/1512.03385}, + eprint = {1512.03385}, + archiveprefix = {arxiv}, + keywords = {/unread,⛔ No DOI found} +} + +@article{heitmannCoyoteUniversePrecision2010, + title = {The {{Coyote Universe}}. {{I}}. {{Precision Determination}} of the {{Nonlinear Matter Power Spectrum}}}, + author = {Heitmann, Katrin and White, Martin and Wagner, Christian and Habib, Salman and Higdon, David}, + year = {2010}, + month = may, + journal = {The Astrophysical Journal}, + volume = {715}, + pages = {104--121}, + issn = {0004-637X}, + doi = {10.1088/0004-637X/715/1/104}, + urldate = {2023-09-25}, + abstract = {Near-future cosmological observations targeted at investigations of dark energy pose stringent requirements on the accuracy of theoretical predictions for the nonlinear clustering of matter. Currently, N-body simulations comprise the only viable approach to this problem. In this paper, we study various sources of computational error and methods to control them. By applying our methodology to a large suite of cosmological simulations we show that results for the (gravity-only) nonlinear matter power spectrum can be obtained at 1\% accuracy out to k \textasciitilde{} 1 h Mpc-1. The key components of these high accuracy simulations are precise initial conditions, very large simulation volumes, sufficient mass resolution, and accurate time stepping. This paper is the first in a series of three; the final aim is a high-accuracy prediction scheme for the nonlinear matter power spectrum that improves current fitting formulae by an order of magnitude.}, + keywords = {/unread,Astrophysics,large-scale structure of universe,methods: numerical}, + annotation = {ADS Bibcode: 2010ApJ...715..104H}, + file = {/Users/nord/Zotero/storage/PYKJ74PX/Heitmann et al. - 2010 - The Coyote Universe. I. Precision Determination of.pdf} +} + +@misc{hlozekResultsPhotometricLSST2020a, + title = {Results of the {{Photometric LSST Astronomical Time-series Classification Challenge}} ({{PLAsTiCC}})}, + author = {Hlo{\v z}ek, R. and Ponder, K. A. and Malz, A. I. and Dai, M. and Narayan, G. and Ishida, E. E. O. and Allam, T. and Bahmanyar, A. and Biswas, R. and Galbany, L. and Jha, S. W. and Jones, D. O. and Kessler, R. and Lochner, M. and Mahabal, A. A. and Mandel, K. S. and {Mart{\'i}nez-Galarza}, J. R. and McEwen, J. D. and Muthukrishna, D. and Peiris, H. V. and Peters, C. M. and Setzer, C. N.}, + year = {2020}, + publisher = {{arXiv}}, + doi = {10.48550/ARXIV.2012.12392}, + copyright = {Creative Commons Attribution 4.0 International}, + keywords = {/unread,Cosmology and Nongalactic Astrophysics (astro-ph.CO),FOS: Physical sciences,High Energy Astrophysical Phenomena (astro-ph.HE),Instrumentation and Methods for Astrophysics (astro-ph.IM)} +} + +@article{huchraSurveyGalaxyRedshifts1983, + title = {A Survey of Galaxy Redshifts. {{IV}} - {{The}} Data}, + author = {Huchra, J. and Davis, M. and Latham, D. and Tonry, J.}, + year = {1983}, + month = jun, + journal = {The Astrophysical Journal Supplement Series}, + volume = {52}, + pages = {89--119}, + issn = {0067-0049}, + doi = {10.1086/190860}, + urldate = {2023-09-25}, + abstract = {We present here the complete list of the best available radial velocities for the 2401 galaxies in the merged Zwicky-Nilson catalog brighter than 14.5 mZ and with bII above +40\textdegree{} or below -30\textdegree. Almost 60\% of the redshifts are from the CfA survey and are accurate to typically 35 km s-1.}, + keywords = {/unread,Astronomical Catalogs,Astronomical Spectroscopy,Astronomy,Galactic Rotation,Luminosity,Radial Velocity,Red Shift}, + annotation = {ADS Bibcode: 1983ApJS...52...89H}, + file = {/Users/nord/Zotero/storage/SQRECNFK/Huchra et al. - 1983 - A survey of galaxy redshifts. IV - The data.pdf} +} + +@article{hunterMatplotlib2DGraphics2007b, + title = {Matplotlib: {{A 2D}} Graphics Environment}, + author = {Hunter, J. D.}, + year = {2007}, + journal = {Computing in Science \& Engineering}, + volume = {9}, + number = {3}, + pages = {90--95}, + publisher = {{IEEE COMPUTER SOC}}, + doi = {10.1109/MCSE.2007.55}, + abstract = {Matplotlib is a 2D graphics package used for Python for application development, interactive scripting, and publication-quality image generation across user interfaces and operating systems.}, + keywords = {/unread} +} + +@article{klypinDARKMATTERHALOS2011, + title = {{{DARK MATTER HALOS IN THE STANDARD COSMOLOGICAL MODEL}}: {{RESULTS FROM THE BOLSHOI SIMULATION}}}, + shorttitle = {{{DARK MATTER HALOS IN THE STANDARD COSMOLOGICAL MODEL}}}, + author = {Klypin, Anatoly A. and {Trujillo-Gomez}, Sebastian and Primack, Joel}, + year = {2011}, + month = oct, + journal = {ApJ}, + volume = {740}, + number = {2}, + pages = {102}, + publisher = {{The American Astronomical Society}}, + issn = {0004-637X}, + doi = {10.1088/0004-637X/740/2/102}, + urldate = {2023-09-25}, + abstract = {Lambda Cold Dark Matter ({$\Lambda$}CDM) is now the standard theory of structure formation in the universe. We present the first results from the new Bolshoi dissipationless cosmological {$\Lambda$}CDM simulation that uses cosmological parameters favored by current observations. The Bolshoi simulation was run in a volume 250 h-1 Mpc on a side using {$\sim$}8 billion particles with mass and force resolution adequate to follow subhalos down to the completeness limit of Vcirc = 50 km s-1 maximum circular velocity. Using merger trees derived from analysis of 180 stored time steps we find the circular velocities of satellites before they fall into their host halos. Using excellent statistics of halos and subhalos ({$\sim$}10 million at every moment and {$\sim$}50 million over the whole history) we present accurate approximations for statistics such as the halo mass function, the concentrations for distinct halos and subhalos, the abundance of halos as a function of their circular velocity, and the abundance and the spatial distribution of subhalos. We find that at high redshifts the concentration falls to a minimum value of about 4.0 and then rises for higher values of halo mass\textemdash a new result. We present approximations for the velocity and mass functions of distinct halos as a function of redshift. We find that while the Sheth\textendash Tormen (ST) approximation for the mass function of halos found by spherical overdensity is quite accurate at low redshifts, the ST formula overpredicts the abundance of halos by nearly an order of magnitude by z = 10. We find that the number of subhalos scales with the circular velocity of the host halo as V1/2host, and that subhalos have nearly the same radial distribution as dark matter particles at radii 0.3\textendash 2 times the host halo virial radius. The subhalo velocity function N({$>$} Vsub) scales as V-3circ. Combining the results of Bolshoi and Via Lactea-II simulations, we find that inside the virial radius of halos with the number of satellites is N({$>$} Vsub) = (Vsub/58 km s-1)-3 for satellite circular velocities in the range 4 km s-1 {$<$} Vsub {$<$} 150 km s-1.}, + langid = {english}, + keywords = {/unread}, + file = {/Users/nord/Zotero/storage/ZVWSMCPK/Klypin et al. - 2011 - DARK MATTER HALOS IN THE STANDARD COSMOLOGICAL MOD.pdf} +} + +@article{krizhevskyCIFAR10CanadianInstitute2017a, + title = {{{CIFAR-10}} ({{Canadian Institute}} for {{Advanced Research}})}, + author = {Krizhevsky, Alex and Nair, Vinod and Hinton, Geoffrey}, + year = {2017}, + month = apr, + abstract = {The CIFAR-10 dataset consists of 60000 32x32 colour images in 10 classes, with 6000 images per class. There are 50000 training images and 10000 test images. The dataset is divided into five training batches and one test batch, each with 10000 images. The test batch contains exactly 1000 randomly-selected images from each class. The training batches contain the remaining images in random order, but some training batches may contain more images from one class than another. Between them, the training batches contain exactly 5000 images from each class.}, + keywords = {/unread,⛔ No DOI found,Dataset} +} + +@inproceedings{lecunMnistDatabaseHandwritten2005, + title = {The Mnist Database of Handwritten Digits}, + author = {LeCun, Yann and Cortes, Corinna}, + year = {2005}, + urldate = {2023-08-10}, + abstract = {Disclosed is an improved articulated bar flail having shearing edges for efficiently shredding materials. An improved shredder cylinder is disclosed with a plurality of these flails circumferentially spaced and pivotally attached to the periphery of a rotatable shaft. Also disclosed is an improved shredder apparatus which has a pair of these shredder cylinders mounted to rotate about spaced parallel axes which cooperates with a conveyer apparatus which has a pair of inclined converging conveyer belts with one of the belts mounted to move with respect to the other belt to allow the transport of articles of various sizes therethrough.}, + keywords = {/unread,⛔ No DOI found}, + file = {/Users/nord/Zotero/storage/WXSAPBPZ/LeCun and Cortes - 2005 - The mnist database of handwritten digits.pdf} +} + +@article{lewisEfficientComputationCMB2000, + title = {Efficient {{Computation}} of {{CMB}} Anisotropies in Closed {{FRW}} Models}, + author = {Lewis, Antony and Challinor, Anthony and Lasenby, Anthony}, + year = {2000}, + month = aug, + journal = {ApJ}, + volume = {538}, + number = {2}, + eprint = {astro-ph/9911177}, + pages = {473--476}, + issn = {0004-637X, 1538-4357}, + doi = {10.1086/309179}, + urldate = {2023-09-25}, + abstract = {We implement the efficient line of sight method to calculate the anisotropy and polarization of the cosmic microwave background for scalar and tensor modes in almost-Friedmann-Robertson-Walker models with positive spatial curvature. We present new results for the polarization power spectra in such models.}, + archiveprefix = {arxiv}, + keywords = {/unread,Astrophysics}, + file = {/Users/nord/Zotero/storage/XUN2S9XC/Lewis et al. - 2000 - Efficient Computation of CMB anisotropies in close.pdf;/Users/nord/Zotero/storage/3DAS4EAI/9911177.html} +} + +@article{manciniBayesianModelComparison2022, + title = {Bayesian Model Comparison for Simulation-Based Inference}, + author = {Mancini, A Spurio and Docherty, M M and Price, M A and McEwen, J D}, + year = {2022}, + abstract = {Comparison of appropriate models to describe observational data is a fundamental task of science. The Bayesian model evidence, or marginal likelihood, is a computationally challenging, yet crucial, quantity to estimate to perform Bayesian model comparison. We introduce a methodology to compute the Bayesian model evidence in simulation-based inference (SBI) scenarios (also often called likelihood-free inference). In particular, we leverage the recently proposed learnt harmonic mean estimator and exploit the fact that it is decoupled from the method used to generate posterior samples, i.e. it requires posterior samples only, which may be generated by any approach. This flexibility, which is lacking in many alternative methods for computing the model evidence, allows us to develop SBI model comparison techniques for the three main neural density estimation approaches, including neural posterior estimation (NPE), neural likelihood estimation (NLE), and neural ratio estimation (NRE). We demonstrate and validate our SBI evidence calculation techniques on a range of inference problems, including a gravitational wave example. Moreover, we further validate the accuracy of the learnt harmonic mean estimator, implemented in the harmonic software, in likelihood-based settings. These results highlight the potential of harmonic as a sampler-agnostic method to estimate the model evidence in both likelihood-based and simulation-based scenarios.}, + langid = {english}, + keywords = {/unread,⛔ No DOI found}, + file = {/Users/nord/Zotero/storage/9KX7FY37/Mancini et al. - 2022 - Bayesian model comparison for simulation-based inf.pdf} +} + +@article{metcalfStrongGravitationalLens2019c, + title = {The Strong Gravitational Lens Finding Challenge}, + author = {Metcalf, R. B. and Meneghetti, M. and Avestruz, C. and Bellagamba, F. and Bom, C. R. and Bertin, E. and Cabanac, R. and Courbin, F. and Davies, A. and {re}, E. Decenci{\`e} and Flamary, R. and Gavazzi, R. and Geiger, M. and Hartley, P. and {Huertas-Company}, M. and Jackson, N. and Jacobs, C. and Jullo, E. and Kneib, J.-P. and Koopmans, L. V. E. and Lanusse, F. and Li, C.-L. and Ma, Q. and Makler, M. and Li, N. and Lightman, M. and Petrillo, C. E. and Serjeant, S. and Sch{\"a}fer, C. and Sonnenfeld, A. and Tagore, A. and Tortora, C. and Tuccillo, D. and Valent{\'i}n, M. B. and {Velasco-Forero}, S. and Kleijn, G. A. Verdoes and Vernardos, G.}, + year = {2019}, + month = may, + journal = {Astronomy \&\textbackslash mathsemicolon Astrophysics}, + volume = {625}, + pages = {A119}, + publisher = {{EDP Sciences}}, + doi = {10.1051/0004-6361/201832797}, + keywords = {/unread} +} + +@article{morganDeeplenstronomyDatasetSimulation2021a, + title = {Deeplenstronomy: {{A}} Dataset Simulation Package for Strong Gravitational Lensing}, + author = {Morgan, Robert and Nord, Brian and Birrer, Simon and Lin, Joshua Yao-Yu and Poh, Jason}, year = {2021}, - publisher = {The Open Journal}, + journal = {Journal of Open Source Software}, volume = {6}, number = {58}, pages = {2854}, - author = {Robert Morgan and Brian Nord and Simon Birrer and Joshua Yao-Yu Lin and Jason Poh}, - title = {deeplenstronomy: A dataset simulation package for strong gravitational lensing}, - journal = {Journal of Open Source Software} -} - -@Article{Hunter:2007, - Author = {Hunter, J. D.}, - Title = {Matplotlib: A 2D graphics environment}, - Journal = {Computing in Science \& Engineering}, - Volume = {9}, - Number = {3}, - Pages = {90--95}, - abstract = {Matplotlib is a 2D graphics package used for Python for - application development, interactive scripting, and publication-quality - image generation across user interfaces and operating systems.}, - publisher = {IEEE COMPUTER SOC}, - doi = {10.1109/MCSE.2007.55}, - year = 2007 - } - -@article{DBLP:journals/corr/WuTWSDH16, - author = {Qi Wu and - Damien Teney and - Peng Wang and - Chunhua Shen and - Anthony R. Dick and - Anton van den Hengel}, - title = {Visual Question Answering: {A} Survey of Methods and Datasets}, - journal = {CoRR}, - volume = {abs/1607.05910}, - year = {2016}, - url = {http://arxiv.org/abs/1607.05910}, - eprinttype = {arXiv}, - eprint = {1607.05910}, - timestamp = {Tue, 19 Mar 2019 13:03:53 +0100}, - biburl = {https://dblp.org/rec/journals/corr/WuTWSDH16.bib}, - bibsource = {dblp computer science bibliography, https://dblp.org} -} - -@misc{https://doi.org/10.48550/arxiv.1809.01691, - doi = {10.48550/ARXIV.1809.01691}, - url = {https://arxiv.org/abs/1809.01691}, - author = {González, Roberto E. and Muñoz, Roberto P. and Hernández, Cristian A.}, - keywords = {Instrumentation and Methods for Astrophysics (astro-ph.IM), FOS: Physical sciences, FOS: Physical sciences}, - title = {Galaxy detection and identification using deep learning and data augmentation}, - publisher = {arXiv}, - year = {2018}, - copyright = {arXiv.org perpetual, non-exclusive license} -} - -@article{ refId0, - author = {{The Astropy Collaboration} and {Robitaille, Thomas P.} and {Tollerud, Erik J.} and {Greenfield, Perry} and {Droettboom, Michael} and {Bray, Erik} and {Aldcroft, Tom} and {Davis, Matt} and {Ginsburg, Adam} and {Price-Whelan, Adrian M.} and {Kerzendorf, Wolfgang E.} and {Conley, Alexander} and {Crighton, Neil} and {Barbary, Kyle} and {Muna, Demitri} and {Ferguson, Henry} and {Grollier, Fr\'ed\'eric} and {Parikh, Madhura M.} and {Nair, Prasanth H.} and {G\"unther, Hans M.} and {Deil, Christoph} and {Woillez, Julien} and {Conseil, Simon} and {Kramer, Roban} and {Turner, James E. H.} and {Singer, Leo} and {Fox, Ryan} and {Weaver, Benjamin A.} and {Zabalza, Victor} and {Edwards, Zachary I.} and {Azalee Bostroem, K.} and {Burke, D. J.} and {Casey, Andrew R.} and {Crawford, Steven M.} and {Dencheva, Nadia} and {Ely, Justin} and {Jenness, Tim} and {Labrie, Kathleen} and {Lim, Pey Lian} and {Pierfederici, Francesco} and {Pontzen, Andrew} and {Ptak, Andy} and {Refsdal, Brian} and {Servillat, Mathieu} and {Streicher, Ole}}, - title = {Astropy: A community Python package for astronomy}, - DOI= "10.1051/0004-6361/201322068", - url= "https://doi.org/10.1051/0004-6361/201322068", - journal = {A\&A}, - year = 2013, - volume = 558, - pages = "A33", - month = "", -} - -@article{astropy:2013, -Adsnote = {Provided by the SAO/NASA Astrophysics Data System}, -Adsurl = {http://adsabs.harvard.edu/abs/2013A%26A...558A..33A}, -Archiveprefix = {arXiv}, -Author = {{Astropy Collaboration} and {Robitaille}, T.~P. and {Tollerud}, E.~J. and {Greenfield}, P. and {Droettboom}, M. and {Bray}, E. and {Aldcroft}, T. and {Davis}, M. and {Ginsburg}, A. and {Price-Whelan}, A.~M. and {Kerzendorf}, W.~E. and {Conley}, A. and {Crighton}, N. and {Barbary}, K. and {Muna}, D. and {Ferguson}, H. and {Grollier}, F. and {Parikh}, M.~M. and {Nair}, P.~H. and {Unther}, H.~M. and {Deil}, C. and {Woillez}, J. and {Conseil}, S. and {Kramer}, R. and {Turner}, J.~E.~H. and {Singer}, L. and {Fox}, R. and {Weaver}, B.~A. and {Zabalza}, V. and {Edwards}, Z.~I. and {Azalee Bostroem}, K. and {Burke}, D.~J. and {Casey}, A.~R. and {Crawford}, S.~M. and {Dencheva}, N. and {Ely}, J. and {Jenness}, T. and {Labrie}, K. and {Lim}, P.~L. and {Pierfederici}, F. and {Pontzen}, A. and {Ptak}, A. and {Refsdal}, B. and {Servillat}, M. and {Streicher}, O.}, -Doi = {10.1051/0004-6361/201322068}, -Eid = {A33}, -Eprint = {1307.6212}, -Journal = {\aap}, -Keywords = {methods: data analysis, methods: miscellaneous, virtual observatory tools}, -Month = oct, -Pages = {A33}, -Primaryclass = {astro-ph.IM}, -Title = {{Astropy: A community Python package for astronomy}}, -Volume = 558, -Year = 2013, -Bdsk-Url-1 = {https://dx.doi.org/10.1051/0004-6361/201322068}} - -@ARTICLE{astropy:2018, - author = {{Astropy Collaboration} and {Price-Whelan}, A.~M. and - {Sip{\H{o}}cz}, B.~M. and {G{\"u}nther}, H.~M. and {Lim}, P.~L. and - {Crawford}, S.~M. and {Conseil}, S. and {Shupe}, D.~L. and - {Craig}, M.~W. and {Dencheva}, N. and {Ginsburg}, A. and {Vand - erPlas}, J.~T. and {Bradley}, L.~D. and {P{\'e}rez-Su{\'a}rez}, D. and - {de Val-Borro}, M. and {Aldcroft}, T.~L. and {Cruz}, K.~L. and - {Robitaille}, T.~P. and {Tollerud}, E.~J. and {Ardelean}, C. and - {Babej}, T. and {Bach}, Y.~P. and {Bachetti}, M. and {Bakanov}, A.~V. and - {Bamford}, S.~P. and {Barentsen}, G. and {Barmby}, P. and - {Baumbach}, A. and {Berry}, K.~L. and {Biscani}, F. and {Boquien}, M. and - {Bostroem}, K.~A. and {Bouma}, L.~G. and {Brammer}, G.~B. and - {Bray}, E.~M. and {Breytenbach}, H. and {Buddelmeijer}, H. and - {Burke}, D.~J. and {Calderone}, G. and {Cano Rodr{\'\i}guez}, J.~L. and - {Cara}, M. and {Cardoso}, J.~V.~M. and {Cheedella}, S. and {Copin}, Y. and - {Corrales}, L. and {Crichton}, D. and {D'Avella}, D. and {Deil}, C. and - {Depagne}, {\'E}. and {Dietrich}, J.~P. and {Donath}, A. and - {Droettboom}, M. and {Earl}, N. and {Erben}, T. and {Fabbro}, S. and - {Ferreira}, L.~A. and {Finethy}, T. and {Fox}, R.~T. and - {Garrison}, L.~H. and {Gibbons}, S.~L.~J. and {Goldstein}, D.~A. and - {Gommers}, R. and {Greco}, J.~P. and {Greenfield}, P. and - {Groener}, A.~M. and {Grollier}, F. and {Hagen}, A. and {Hirst}, P. and - {Homeier}, D. and {Horton}, A.~J. and {Hosseinzadeh}, G. and {Hu}, L. and - {Hunkeler}, J.~S. and {Ivezi{\'c}}, {\v{Z}}. and {Jain}, A. and - {Jenness}, T. and {Kanarek}, G. and {Kendrew}, S. and {Kern}, N.~S. and - {Kerzendorf}, W.~E. and {Khvalko}, A. and {King}, J. and {Kirkby}, D. and - {Kulkarni}, A.~M. and {Kumar}, A. and {Lee}, A. and {Lenz}, D. and - {Littlefair}, S.~P. and {Ma}, Z. and {Macleod}, D.~M. and - {Mastropietro}, M. and {McCully}, C. and {Montagnac}, S. and - {Morris}, B.~M. and {Mueller}, M. and {Mumford}, S.~J. and {Muna}, D. and - {Murphy}, N.~A. and {Nelson}, S. and {Nguyen}, G.~H. and - {Ninan}, J.~P. and {N{\"o}the}, M. and {Ogaz}, S. and {Oh}, S. and - {Parejko}, J.~K. and {Parley}, N. and {Pascual}, S. and {Patil}, R. and - {Patil}, A.~A. and {Plunkett}, A.~L. and {Prochaska}, J.~X. and - {Rastogi}, T. and {Reddy Janga}, V. and {Sabater}, J. and - {Sakurikar}, P. and {Seifert}, M. and {Sherbert}, L.~E. and - {Sherwood-Taylor}, H. and {Shih}, A.~Y. and {Sick}, J. and - {Silbiger}, M.~T. and {Singanamalla}, S. and {Singer}, L.~P. and - {Sladen}, P.~H. and {Sooley}, K.~A. and {Sornarajah}, S. and - {Streicher}, O. and {Teuben}, P. and {Thomas}, S.~W. and - {Tremblay}, G.~R. and {Turner}, J.~E.~H. and {Terr{\'o}n}, V. and - {van Kerkwijk}, M.~H. and {de la Vega}, A. and {Watkins}, L.~L. and - {Weaver}, B.~A. and {Whitmore}, J.~B. and {Woillez}, J. and - {Zabalza}, V. and {Astropy Contributors}}, - title = "{The Astropy Project: Building an Open-science Project and Status of the v2.0 Core Package}", - journal = {\aj}, - keywords = {methods: data analysis, methods: miscellaneous, methods: statistical, reference systems, Astrophysics - Instrumentation and Methods for Astrophysics}, - year = 2018, - month = sep, - volume = {156}, - number = {3}, - eid = {123}, - pages = {123}, - doi = {10.3847/1538-3881/aabc4f}, -archivePrefix = {arXiv}, - eprint = {1801.02634}, - primaryClass = {astro-ph.IM}, - adsurl = {https://ui.adsabs.harvard.edu/abs/2018AJ....156..123A}, - adsnote = {Provided by the SAO/NASA Astrophysics Data System} -} - - - -@article{DBLP:journals/corr/HeZRS15, - author = {Kaiming He and - Xiangyu Zhang and - Shaoqing Ren and - Jian Sun}, - title = {Deep Residual Learning for Image Recognition}, - journal = {CoRR}, - volume = {abs/1512.03385}, - year = {2015}, - url = {http://arxiv.org/abs/1512.03385}, - eprinttype = {arXiv}, - eprint = {1512.03385}, - timestamp = {Wed, 17 Apr 2019 17:23:45 +0200}, - biburl = {https://dblp.org/rec/journals/corr/HeZRS15.bib}, - bibsource = {dblp computer science bibliography, https://dblp.org} -} - -@incollection{NEURIPS2019_9015, -title = {PyTorch: An Imperative Style, High-Performance Deep Learning Library}, -author = {Paszke, Adam and Gross, Sam and Massa, Francisco and Lerer, Adam and Bradbury, James and Chanan, Gregory and Killeen, Trevor and Lin, Zeming and Gimelshein, Natalia and Antiga, Luca and Desmaison, Alban and Kopf, Andreas and Yang, Edward and DeVito, Zachary and Raison, Martin and Tejani, Alykhan and Chilamkurthy, Sasank and Steiner, Benoit and Fang, Lu and Bai, Junjie and Chintala, Soumith}, -booktitle = {Advances in Neural Information Processing Systems 32}, -editor = {H. Wallach and H. Larochelle and A. Beygelzimer and F. d\textquotesingle Alch\'{e}-Buc and E. Fox and R. Garnett}, -pages = {8024--8035}, -year = {2019}, -publisher = {Curran Associates, Inc.}, -url = {http://papers.neurips.cc/paper/9015-pytorch-an-imperative-style-high-performance-deep-learning-library.pdf} -} - -@Article{ harris2020array, + publisher = {{The Open Journal}}, + doi = {10.21105/joss.02854}, + keywords = {/unread} +} + +@incollection{paszkePyTorchImperativeStyle2019b, + title = {{{PyTorch}}: {{An Imperative Style}}, {{High-Performance Deep Learning Library}}}, + booktitle = {Advances in {{Neural Information Processing Systems}} 32}, + author = {Paszke, Adam and Gross, Sam and Massa, Francisco and Lerer, Adam and Bradbury, James and Chanan, Gregory and Killeen, Trevor and Lin, Zeming and Gimelshein, Natalia and Antiga, Luca and Desmaison, Alban and Kopf, Andreas and Yang, Edward and DeVito, Zachary and Raison, Martin and Tejani, Alykhan and Chilamkurthy, Sasank and Steiner, Benoit and Fang, Lu and Bai, Junjie and Chintala, Soumith}, + editor = {Wallach, H. and Larochelle, H. and Beygelzimer, A. and d'{\aftergroup\ignorespaces} {Alch{\'e}-Buc}, F. and Fox, E. and Garnett, R.}, + year = {2019}, + pages = {8024--8035}, + publisher = {{Curran Associates, Inc.}}, + keywords = {/unread} +} + +@misc{piaGeant4ScientificLiterature2009, + title = {Geant4 in {{Scientific Literature}}}, + author = {Pia, M. G. and Basaglia, T. and Bell, Z. W. and Dressendorfer, P. V.}, + year = {2009}, + month = dec, + number = {arXiv:0912.0360}, + eprint = {0912.0360}, + primaryclass = {physics}, + publisher = {{arXiv}}, + doi = {10.48550/arXiv.0912.0360}, + urldate = {2023-09-25}, + abstract = {The Geant4 reference paper published in Nuclear Instruments and Methods A in 2003 has become the most cited publication in the whole Nuclear Science and Technology category of Thomson-Reuter's Journal Citation Reports. It is currently the second most cited article among the publications authored by two major research institutes, CERN and INFN. An overview of Geant4 presence (and absence) in scholarly literature is presented; the patterns of Geant4 citations are quantitatively examined and discussed.}, + archiveprefix = {arxiv}, + keywords = {/unread,Computer Science - Digital Libraries,Physics - Computational Physics}, + file = {/Users/nord/Zotero/storage/XHUDQSJ7/Pia et al. - 2009 - Geant4 in Scientific Literature.pdf;/Users/nord/Zotero/storage/T9UNU3GH/0912.html} +} + +@article{ringermacherNewFormulaDescribing2009a, + title = {A New Formula Describing the Scaffold Structure of Spiral Galaxies}, + author = {Ringermacher, Harry I. and Mead, Lawrence R.}, + year = {2009}, + month = jul, + journal = {Monthly Notices of the Royal Astronomical Society}, + volume = {397}, + number = {1}, + pages = {164--171}, + publisher = {{Oxford University Press (OUP)}}, + doi = {10.1111/j.1365-2966.2009.14950.x}, + keywords = {/unread} +} + +@misc{roweGalSimModularGalaxy2015, + title = {{{GalSim}}: {{The}} Modular Galaxy Image Simulation Toolkit}, + shorttitle = {{{GalSim}}}, + author = {Rowe, Barnaby and Jarvis, Mike and Mandelbaum, Rachel and Bernstein, Gary M. and Bosch, James and Simet, Melanie and Meyers, Joshua E. and Kacprzak, Tomasz and Nakajima, Reiko and Zuntz, Joe and Miyatake, Hironao and Dietrich, Joerg P. and Armstrong, Robert and Melchior, Peter and Gill, Mandeep S. S.}, + year = {2015}, + month = feb, + number = {arXiv:1407.7676}, + eprint = {1407.7676}, + primaryclass = {astro-ph}, + publisher = {{arXiv}}, + doi = {10.48550/arXiv.1407.7676}, + urldate = {2023-09-14}, + abstract = {GALSIM is a collaborative, open-source project aimed at providing an image simulation tool of enduring benefit to the astronomical community. It provides a software library for generating images of astronomical objects such as stars and galaxies in a variety of ways, efficiently handling image transformations and operations such as convolution and rendering at high precision. We describe the GALSIM software and its capabilities, including necessary theoretical background. We demonstrate that the performance of GALSIM meets the stringent requirements of high precision image analysis applications such as weak gravitational lensing, for current datasets and for the Stage IV dark energy surveys of the Large Synoptic Survey Telescope, ESA's Euclid mission, and NASA's WFIRST-AFTA mission. The GALSIM project repository is public and includes the full code history, all open and closed issues, installation instructions, documentation, and wiki pages (including a Frequently Asked Questions section). The GALSIM repository can be found at https://github.com/GalSim-developers/GalSim .}, + archiveprefix = {arxiv}, + keywords = {/unread,85-04,Astrophysics - Cosmology and Nongalactic Astrophysics,Astrophysics - Instrumentation and Methods for Astrophysics} +} + +@misc{roweGalSimModularGalaxy2015a, + title = {{{GalSim}}: {{The}} Modular Galaxy Image Simulation Toolkit}, + shorttitle = {{{GalSim}}}, + author = {Rowe, Barnaby and Jarvis, Mike and Mandelbaum, Rachel and Bernstein, Gary M. and Bosch, James and Simet, Melanie and Meyers, Joshua E. and Kacprzak, Tomasz and Nakajima, Reiko and Zuntz, Joe and Miyatake, Hironao and Dietrich, Joerg P. and Armstrong, Robert and Melchior, Peter and Gill, Mandeep S. S.}, + year = {2015}, + month = feb, + number = {arXiv:1407.7676}, + eprint = {1407.7676}, + primaryclass = {astro-ph}, + publisher = {{arXiv}}, + doi = {10.48550/arXiv.1407.7676}, + urldate = {2023-09-14}, + abstract = {GALSIM is a collaborative, open-source project aimed at providing an image simulation tool of enduring benefit to the astronomical community. It provides a software library for generating images of astronomical objects such as stars and galaxies in a variety of ways, efficiently handling image transformations and operations such as convolution and rendering at high precision. We describe the GALSIM software and its capabilities, including necessary theoretical background. We demonstrate that the performance of GALSIM meets the stringent requirements of high precision image analysis applications such as weak gravitational lensing, for current datasets and for the Stage IV dark energy surveys of the Large Synoptic Survey Telescope, ESA's Euclid mission, and NASA's WFIRST-AFTA mission. The GALSIM project repository is public and includes the full code history, all open and closed issues, installation instructions, documentation, and wiki pages (including a Frequently Asked Questions section). The GALSIM repository can be found at https://github.com/GalSim-developers/GalSim .}, + archiveprefix = {arxiv}, + keywords = {/unread,85-04,Astrophysics - Cosmology and Nongalactic Astrophysics,Astrophysics - Instrumentation and Methods for Astrophysics} +} + +@misc{russakovskyImageNetLargeScale2015a, + title = {{{ImageNet Large Scale Visual Recognition Challenge}}}, + author = {Russakovsky, Olga and Deng, Jia and Su, Hao and Krause, Jonathan and Satheesh, Sanjeev and Ma, Sean and Huang, Zhiheng and Karpathy, Andrej and Khosla, Aditya and Bernstein, Michael and Berg, Alexander C. and {Fei-Fei}, Li}, + year = {2015}, + month = jan, + number = {arXiv:1409.0575}, + eprint = {1409.0575}, + primaryclass = {cs}, + publisher = {{arXiv}}, + doi = {10.48550/arXiv.1409.0575}, + urldate = {2023-09-25}, + abstract = {The ImageNet Large Scale Visual Recognition Challenge is a benchmark in object category classification and detection on hundreds of object categories and millions of images. The challenge has been run annually from 2010 to present, attracting participation from more than fifty institutions. This paper describes the creation of this benchmark dataset and the advances in object recognition that have been possible as a result. We discuss the challenges of collecting large-scale ground truth annotation, highlight key breakthroughs in categorical object recognition, provide a detailed analysis of the current state of the field of large-scale image classification and object detection, and compare the state-of-the-art computer vision accuracy with human accuracy. We conclude with lessons learned in the five years of the challenge, and propose future directions and improvements.}, + archiveprefix = {arxiv}, + keywords = {/unread,Computer Science - Computer Vision and Pattern Recognition,I.4.8,I.5.2}, + file = {/Users/nord/Zotero/storage/9SIIIV3I/Russakovsky et al. - 2015 - ImageNet Large Scale Visual Recognition Challenge.pdf;/Users/nord/Zotero/storage/H96RZZ6R/1409.html} +} + +@article{safonovaRosellaMockCatalogue2021, + title = {Rosella: {{A}} Mock Catalogue from the {{P-Millennium}} Simulation}, + shorttitle = {Rosella}, + author = {Safonova, Sasha and Norberg, Peder and Cole, Shaun}, + year = {2021}, + month = may, + journal = {Monthly Notices of the Royal Astronomical Society}, + volume = {505}, + number = {1}, + eprint = {2009.00005}, + primaryclass = {astro-ph}, + pages = {325--338}, + issn = {0035-8711, 1365-2966}, + doi = {10.1093/mnras/stab1286}, + urldate = {2023-08-14}, + abstract = {The scientific exploitation of the Dark Energy Spectroscopic Instrument Bright Galaxy Survey (DESI BGS) data requires the construction of mocks with galaxy population properties closely mimicking those of the actual DESI BGS targets. We create a high fidelity mock galaxy catalogue, including information about galaxies and their host dark matter subhaloes. The mock catalogue uses subhalo abundance matching (SHAM) with scatter to populate the P-Millennium N-body simulation with galaxies at the median BGS redshift of \textasciitilde{} 0.2, using formation redshift information to assign (g-r) rest-frame colours. The mock provides information about r-band absolute magnitudes, (g-r) rest-frame colours, 3D positions and velocities of a complete sample of DESI BGS galaxies in a volume of (542 Mpc/h)\^3, as well as the masses of host dark matter haloes. This P-Millennium DESI BGS mock catalogue is ideally suited for the tuning of approximate mocks unable to resolve subhaloes that DESI BGS galaxies reside in, to test for systematics in analysis pipelines and to interpret (non-cosmological focused) DESI BGS analysis.}, + archiveprefix = {arxiv}, + keywords = {/unread,Astrophysics - Cosmology and Nongalactic Astrophysics}, + file = {/Users/nord/Zotero/storage/P5D533XG/Safonova et al. - 2021 - Rosella A mock catalogue from the P-Millennium si.pdf;/Users/nord/Zotero/storage/5BMHQKZN/2009.html} +} + +@article{schayeEAGLEProjectSimulating2015, + title = {The {{EAGLE}} Project: Simulating the Evolution and Assembly of Galaxies and Their Environments}, + shorttitle = {The {{EAGLE}} Project}, + author = {Schaye, Joop and Crain, Robert A. and Bower, Richard G. and Furlong, Michelle and Schaller, Matthieu and Theuns, Tom and Dalla Vecchia, Claudio and Frenk, Carlos S. and McCarthy, I. G. and Helly, John C. and Jenkins, Adrian and {Rosas-Guevara}, Y. M. and White, Simon D. M. and Baes, Maarten and Booth, C. M. and Camps, Peter and Navarro, Julio F. and Qu, Yan and Rahmati, Alireza and Sawala, Till and Thomas, Peter A. and Trayford, James}, + year = {2015}, + month = jan, + journal = {Monthly Notices of the Royal Astronomical Society}, + volume = {446}, + pages = {521--554}, + issn = {0035-8711}, + doi = {10.1093/mnras/stu2058}, + urldate = {2023-09-25}, + abstract = {We introduce the Virgo Consortium's Evolution and Assembly of GaLaxies and their Environments (EAGLE) project, a suite of hydrodynamical simulations that follow the formation of galaxies and supermassive black holes in cosmologically representative volumes of a standard {$\Lambda$} cold dark matter universe. We discuss the limitations of such simulations in light of their finite resolution and poorly constrained subgrid physics, and how these affect their predictive power. One major improvement is our treatment of feedback from massive stars and active galactic nuclei (AGN) in which thermal energy is injected into the gas without the need to turn off cooling or decouple hydrodynamical forces, allowing winds to develop without predetermined speed or mass loading factors. Because the feedback efficiencies cannot be predicted from first principles, we calibrate them to the present-day galaxy stellar mass function and the amplitude of the galaxy-central black hole mass relation, also taking galaxy sizes into account. The observed galaxy stellar mass function is reproduced to {$\lessequivlnt$} 0.2 dex over the full resolved mass range, 108 {$<$} M*/M{$\odot$} {$\lessequivlnt$} 1011, a level of agreement close to that attained by semi-analytic models, and unprecedented for hydrodynamical simulations. We compare our results to a representative set of low-redshift observables not considered in the calibration, and find good agreement with the observed galaxy specific star formation rates, passive fractions, Tully-Fisher relation, total stellar luminosities of galaxy clusters, and column density distributions of intergalactic C IV and O VI. While the mass-metallicity relations for gas and stars are consistent with observations for M* {$\greaterequivlnt$} 109 M{$\odot$} (M* {$\greaterequivlnt$} 1010 M{$\odot$} at intermediate resolution), they are insufficiently steep at lower masses. For the reference model, the gas fractions and temperatures are too high for clusters of galaxies, but for galaxy groups these discrepancies can be resolved by adopting a higher heating temperature in the subgrid prescription for AGN feedback. The EAGLE simulation suite, which also includes physics variations and higher resolution zoomed-in volumes described elsewhere, constitutes a valuable new resource for studies of galaxy formation.}, + keywords = {/unread,Astrophysics - Astrophysics of Galaxies,Astrophysics - Cosmology and Nongalactic Astrophysics,cosmology: theory,galaxies: evolution,galaxies: formation,methods: numerical}, + annotation = {ADS Bibcode: 2015MNRAS.446..521S}, + file = {/Users/nord/Zotero/storage/TVPFJY5R/Schaye et al. - 2015 - The EAGLE project simulating the evolution and as.pdf} +} + +@article{sjostrandPYTHIAEventGenerator2020, + title = {The {{PYTHIA Event Generator}}: {{Past}}, {{Present}} and {{Future}}}, + shorttitle = {The {{PYTHIA Event Generator}}}, + author = {Sj{\"o}strand, Torbj{\"o}rn}, + year = {2020}, + month = jan, + journal = {Computer Physics Communications}, + volume = {246}, + eprint = {1907.09874}, + primaryclass = {hep-ph, physics:physics}, + pages = {106910}, + issn = {00104655}, + doi = {10.1016/j.cpc.2019.106910}, + urldate = {2023-09-25}, + abstract = {The evolution of the widely-used PYTHIA particle physics event generator is outlined, from the early days to the current status and plans. The key decisions and the development of the major physics components are put in context.}, + archiveprefix = {arxiv}, + keywords = {/unread,High Energy Physics - Phenomenology,Physics - History and Philosophy of Physics}, + file = {/Users/nord/Zotero/storage/YSFJYQBP/Sjöstrand - 2020 - The PYTHIA Event Generator Past, Present and Futu.pdf;/Users/nord/Zotero/storage/ALTVEJMZ/1907.html} +} + +@misc{SOXSSimulatedObservations, + title = {{{SOXS}}: {{Simulated Observations}} of {{X-ray Sources}} \textemdash{} {{SOXS}} 4.6.0 Documentation}, + urldate = {2023-09-25}, + howpublished = {http://hea-www.cfa.harvard.edu/soxs/index.html}, + keywords = {/unread} +} + +@article{springelCosmologicalSimulationCode2005, + title = {The Cosmological Simulation Code {{GADGET-2}}}, + author = {Springel, Volker}, + year = {2005}, + month = dec, + journal = {Monthly Notices of the Royal Astronomical Society}, + volume = {364}, + pages = {1105--1134}, + issn = {0035-8711}, + doi = {10.1111/j.1365-2966.2005.09655.x}, + urldate = {2023-09-25}, + abstract = {We discuss the cosmological simulation code GADGET-2, a new massively parallel TreeSPH code, capable of following a collisionless fluid with the N-body method, and an ideal gas by means of smoothed particle hydrodynamics (SPH). Our implementation of SPH manifestly conserves energy and entropy in regions free of dissipation, while allowing for fully adaptive smoothing lengths. Gravitational forces are computed with a hierarchical multipole expansion, which can optionally be applied in the form of a TreePM algorithm, where only short-range forces are computed with the `tree' method while long-range forces are determined with Fourier techniques. Time integration is based on a quasi-symplectic scheme where long-range and short-range forces can be integrated with different time-steps. Individual and adaptive short-range time-steps may also be employed. The domain decomposition used in the parallelization algorithm is based on a space-filling curve, resulting in high flexibility and tree force errors that do not depend on the way the domains are cut. The code is efficient in terms of memory consumption and required communication bandwidth. It has been used to compute the first cosmological N-body simulation with more than 1010 dark matter particles, reaching a homogeneous spatial dynamic range of 105 per dimension in a three-dimensional box. It has also been used to carry out very large cosmological SPH simulations that account for radiative cooling and star formation, reaching total particle numbers of more than 250 million. We present the algorithms used by the code and discuss their accuracy and performance using a number of test problems. GADGET-2 is publicly released to the research community.}, + keywords = {/unread,Astrophysics,dark matter,galaxies: interactions,methods: numerical}, + annotation = {ADS Bibcode: 2005MNRAS.364.1105S}, + file = {/Users/nord/Zotero/storage/J74X89TK/Springel - 2005 - The cosmological simulation code GADGET-2.pdf} +} + +@article{springelSimulationsFormationEvolution2005, + title = {Simulations of the Formation, Evolution and Clustering of Galaxies and Quasars}, + author = {Springel, Volker and White, Simon D. M. and Jenkins, Adrian and Frenk, Carlos S. and Yoshida, Naoki and Gao, Liang and Navarro, Julio and Thacker, Robert and Croton, Darren and Helly, John and Peacock, John A. and Cole, Shaun and Thomas, Peter and Couchman, Hugh and Evrard, August and Colberg, J{\"o}rg and Pearce, Frazer}, + year = {2005}, + month = jun, + journal = {Nature}, + volume = {435}, + pages = {629--636}, + issn = {0028-0836}, + doi = {10.1038/nature03597}, + urldate = {2023-09-25}, + abstract = {The cold dark matter model has become the leading theoretical picture for the formation of structure in the Universe. This model, together with the theory of cosmic inflation, makes a clear prediction for the initial conditions for structure formation and predicts that structures grow hierarchically through gravitational instability. Testing this model requires that the precise measurements delivered by galaxy surveys can be compared to robust and equally precise theoretical calculations. Here we present a simulation of the growth of dark matter structure using 2,1603 particles, following them from redshift z = 127 to the present in a cube-shaped region 2.230 billion lightyears on a side. In postprocessing, we also follow the formation and evolution of the galaxies and quasars. We show that baryon-induced features in the initial conditions of the Universe are reflected in distorted form in the low-redshift galaxy distribution, an effect that can be used to constrain the nature of dark energy with future generations of observational surveys of galaxies.}, + keywords = {/unread,Astrophysics}, + annotation = {ADS Bibcode: 2005Natur.435..629S} +} + +@article{theastropycollaborationAstropyCommunityPython2013a, + title = {Astropy: {{A}} Community {{Python}} Package for Astronomy}, + author = {{The Astropy Collaboration} and {Robitaille, Thomas P.} and {Tollerud, Erik J.} and {Greenfield, Perry} and {Droettboom, Michael} and {Bray, Erik} and {Aldcroft, Tom} and {Davis, Matt} and {Ginsburg, Adam} and {Price-Whelan, Adrian M.} and {Kerzendorf, Wolfgang E.} and {Conley, Alexander} and {Crighton, Neil} and {Barbary, Kyle} and {Muna, Demitri} and {Ferguson, Henry} and {Grollier, Fr\'ed\'eric} and {Parikh, Madhura M.} and {Nair, Prasanth H.} and {G\"unther, Hans M.} and {Deil, Christoph} and {Woillez, Julien} and {Conseil, Simon} and {Kramer, Roban} and {Turner, James E. H.} and {Singer, Leo} and {Fox, Ryan} and {Weaver, Benjamin A.} and {Zabalza, Victor} and {Edwards, Zachary I.} and {Azalee Bostroem, K.} and {Burke, D. J.} and {Casey, Andrew R.} and {Crawford, Steven M.} and {Dencheva, Nadia} and {Ely, Justin} and {Jenness, Tim} and {Labrie, Kathleen} and {Lim, Pey Lian} and {Pierfederici, Francesco} and {Pontzen, Andrew} and {Ptak, Andy} and {Refsdal, Brian} and {Servillat, Mathieu} and {Streicher, Ole}}, + year = {2013}, + journal = {A\&A}, + volume = {558}, + pages = {A33}, + doi = {10.1051/0004-6361/201322068}, + keywords = {/unread} +} + +@article{villaescusa-navarroCAMELSProjectCosmology2021, + title = {The {{CAMELS Project}}: {{Cosmology}} and {{Astrophysics}} with {{Machine-learning Simulations}}}, + shorttitle = {The {{CAMELS Project}}}, + author = {{Villaescusa-Navarro}, Francisco and {Angl{\'e}s-Alc{\'a}zar}, Daniel and Genel, Shy and Spergel, David N. and Somerville, Rachel S. and Dave, Romeel and Pillepich, Annalisa and Hernquist, Lars and Nelson, Dylan and Torrey, Paul and Narayanan, Desika and Li, Yin and Philcox, Oliver and Torre, Valentina La and Delgado, Ana Maria and Ho, Shirley and Hassan, Sultan and Burkhart, Blakesley and Wadekar, Digvijay and Battaglia, Nicholas and Contardo, Gabriella and Bryan, Greg L.}, + year = {2021}, + month = jul, + journal = {ApJ}, + volume = {915}, + number = {1}, + pages = {71}, + publisher = {{The American Astronomical Society}}, + issn = {0004-637X}, + doi = {10.3847/1538-4357/abf7ba}, + urldate = {2023-09-25}, + abstract = {We present the Cosmology and Astrophysics with MachinE Learning Simulations (CAMELS) project. CAMELS is a suite of 4233 cosmological simulations of volume each: 2184 state-of-the-art (magneto)hydrodynamic simulations run with the AREPO and GIZMO codes, employing the same baryonic subgrid physics as the IllustrisTNG and SIMBA simulations, and 2049 N-body simulations. The goal of the CAMELS project is to provide theory predictions for different observables as a function of cosmology and astrophysics, and it is the largest suite of cosmological (magneto)hydrodynamic simulations designed to train machine-learning algorithms. CAMELS contains thousands of different cosmological and astrophysical models by way of varying {$\Omega$} m , {$\sigma$} 8, and four parameters controlling stellar and active galactic nucleus feedback, following the evolution of more than 100 billion particles and fluid elements over a combined volume of . We describe the simulations in detail and characterize the large range of conditions represented in terms of the matter power spectrum, cosmic star formation rate density, galaxy stellar mass function, halo baryon fractions, and several galaxy scaling relations. We show that the IllustrisTNG and SIMBA suites produce roughly similar distributions of galaxy properties over the full parameter space but significantly different halo baryon fractions and baryonic effects on the matter power spectrum. This emphasizes the need for marginalizing over baryonic effects to extract the maximum amount of information from cosmological surveys. We illustrate the unique potential of CAMELS using several machine-learning applications, including nonlinear interpolation, parameter estimation, symbolic regression, data generation with Generative Adversarial Networks, dimensionality reduction, and anomaly detection.}, + langid = {english}, + keywords = {/unread}, + file = {/Users/nord/Zotero/storage/IVIE2VNR/Villaescusa-Navarro et al. - 2021 - The CAMELS Project Cosmology and Astrophysics wit.pdf} +} + +@article{villaescusa-navarroQuijoteSimulations2020, + title = {The {{Quijote Simulations}}}, + author = {{Villaescusa-Navarro}, Francisco and Hahn, ChangHoon and Massara, Elena and Banerjee, Arka and Delgado, Ana Maria and Ramanah, Doogesh Kodi and Charnock, Tom and Giusarma, Elena and Li, Yin and Allys, Erwan and Brochard, Antoine and Uhlemann, Cora and Chiang, Chi-Ting and He, Siyu and Pisani, Alice and Obuljen, Andrej and Feng, Yu and Castorina, Emanuele and Contardo, Gabriella and Kreisch, Christina D. and Nicola, Andrina and Alsing, Justin and Scoccimarro, Roman and Verde, Licia and Viel, Matteo and Ho, Shirley and Mallat, Stephane and Wandelt, Benjamin and Spergel, David N.}, + year = {2020}, + month = aug, + journal = {ApJS}, + volume = {250}, + number = {1}, + pages = {2}, + publisher = {{The American Astronomical Society}}, + issn = {0067-0049}, + doi = {10.3847/1538-4365/ab9d82}, + urldate = {2023-09-25}, + abstract = {The Quijote simulations are a set of 44,100 full N-body simulations spanning more than 7000 cosmological models in the hyperplane. At a single redshift, the simulations contain more than 8.5 trillion particles over a combined volume of 44,100 each simulation follows the evolution of 2563, 5123, or 10243 particles in a box of 1 h-1 Gpc length. Billions of dark matter halos and cosmic voids have been identified in the simulations, whose runs required more than 35 million core hours. The Quijote simulations have been designed for two main purposes: (1) to quantify the information content on cosmological observables and (2) to provide enough data to train machine-learning algorithms. In this paper, we describe the simulations and show a few of their applications. We also release the petabyte of data generated, comprising hundreds of thousands of simulation snapshots at multiple redshifts; halo and void catalogs; and millions of summary statistics, such as power spectra, bispectra, correlation functions, marked power spectra, and estimated probability density functions.}, + langid = {english}, + keywords = {/unread} +} + +@article{vogelsbergerIntroducingIllustrisProject2014b, + title = {Introducing the {{Illustris Project}}: Simulating the Coevolution of Dark and Visible Matter in the {{Universe}}}, + shorttitle = {Introducing the {{Illustris Project}}}, + author = {Vogelsberger, Mark and Genel, Shy and Springel, Volker and Torrey, Paul and Sijacki, Debora and Xu, Dandan and Snyder, Greg and Nelson, Dylan and Hernquist, Lars}, + year = {2014}, + month = oct, + journal = {Monthly Notices of the Royal Astronomical Society}, + volume = {444}, + number = {2}, + pages = {1518--1547}, + issn = {0035-8711}, + doi = {10.1093/mnras/stu1536}, + urldate = {2023-09-25}, + abstract = {We introduce the Illustris Project, a series of large-scale hydrodynamical simulations of galaxy formation. The highest resolution simulation, Illustris-1, covers a volume of (106.5舁Mpc)3, has a dark mass resolution of 6.26 \texttimes{} 106舁M{$\odot$}, and an initial baryonic matter mass resolution of 1.26 \texttimes{} 106舁M{$\odot$}. At z~=~0 gravitational forces are softened on scales of 710舁pc, and the smallest hydrodynamical gas cells have an extent of 48舁pc. We follow the dynamical evolution of 2~\texttimes ~18203 resolution elements and in addition passively evolve 18203 Monte Carlo tracer particles reaching a total particle count of more than 18 billion. The galaxy formation model includes: primordial and metal-line cooling with self-shielding corrections, stellar evolution, stellar feedback, gas recycling, chemical enrichment, supermassive black hole growth, and feedback from active galactic nuclei. Here we describe the simulation suite, and contrast basic predictions of our model for the present-day galaxy population with observations of the local universe. At z~=~0 our simulation volume contains about 40舁000 well-resolved galaxies covering a diverse range of morphologies and colours including early-type, late-type and irregular galaxies. The simulation reproduces reasonably well the cosmic star formation rate density, the galaxy luminosity function, and baryon conversion efficiency at z~=~0. It also qualitatively captures the impact of galaxy environment on the red fractions of galaxies. The internal velocity structure of selected well-resolved disc galaxies obeys the stellar and baryonic Tully\textendash Fisher relation together with flat circular velocity curves. In the well-resolved regime, the simulation reproduces the observed mix of early-type and late-type galaxies. Our model predicts a halo mass dependent impact of baryonic effects on the halo mass function and the masses of haloes caused by feedback from supernova and active galactic nuclei.}, + keywords = {/unread}, + file = {/Users/nord/Zotero/storage/CFJK54TT/Vogelsberger et al. - 2014 - Introducing the Illustris Project simulating the .pdf} +} + +@misc{WelcomePixellDocumentation, + title = {Welcome to Pixell's Documentation! \textemdash{} Pixell 0.19.2+22.Ga40a155.Dirty Documentation}, + urldate = {2023-09-25}, + howpublished = {https://pixell.readthedocs.io/en/latest/}, + keywords = {/unread}, + file = {/Users/nord/Zotero/storage/TX3HBU4N/latest.html} +} + +@article{willettGalaxyZooDetailed2013b, + title = {Galaxy {{Zoo}} 2: Detailed Morphological Classifications for 304 122 Galaxies from the {{Sloan Digital Sky Survey}}}, + author = {Willett, Kyle W. and Lintott, Chris J. and Bamford, Steven P. and Masters, Karen L. and Simmons, Brooke D. and Casteels, Kevin R. V. and Edmondson, Edward M. and Fortson, Lucy F. and Kaviraj, Sugata and Keel, William C. and Melvin, Thomas and Nichol, Robert C. and Raddick, M. Jordan and Schawinski, Kevin and Simpson, Robert J. and Skibba, Ramin A. and Smith, Arfon M. and Thomas, Daniel}, + year = {2013}, + month = sep, + journal = {Monthly Notices of the Royal Astronomical Society}, + volume = {435}, + number = {4}, + pages = {2835--2860}, + publisher = {{Oxford University Press (OUP)}}, + doi = {10.1093/mnras/stt1458}, + keywords = {/unread} +} + +@article{wuVisualQuestionAnswering2016a, + title = {Visual {{Question Answering}}: {{A Survey}} of {{Methods}} and {{Datasets}}}, + author = {Wu, Qi and Teney, Damien and Wang, Peng and Shen, Chunhua and Dick, Anthony R. and van den Hengel, Anton}, + year = {2016}, + journal = {CoRR}, + volume = {abs/1607.05910}, + eprint = {1607.05910}, + archiveprefix = {arxiv}, + keywords = {/unread,⛔ No DOI found} +} + +@article{yorkSloanDigitalSky2000, + title = {The {{Sloan Digital Sky Survey}}: {{Technical Summary}}}, + shorttitle = {The {{Sloan Digital Sky Survey}}}, + author = {York, Donald G. and Adelman, J. and Anderson, Jr., John E. and Anderson, Scott F. and Annis, James and Bahcall, Neta A. and Bakken, J. A. and Barkhouser, Robert and Bastian, Steven and Berman, Eileen and Boroski, William N. and Bracker, Steve and Briegel, Charlie and Briggs, John W. and Brinkmann, J. and Brunner, Robert and Burles, Scott and Carey, Larry and Carr, Michael A. and Castander, Francisco J. and Chen, Bing and Colestock, Patrick L. and Connolly, A. J. and Crocker, J. H. and Csabai, Istv{\'a}n and Czarapata, Paul C. and Davis, John Eric and Doi, Mamoru and Dombeck, Tom and Eisenstein, Daniel and Ellman, Nancy and Elms, Brian R. and Evans, Michael L. and Fan, Xiaohui and Federwitz, Glenn R. and Fiscelli, Larry and Friedman, Scott and Frieman, Joshua A. and Fukugita, Masataka and Gillespie, Bruce and Gunn, James E. and Gurbani, Vijay K. and {de Haas}, Ernst and Haldeman, Merle and Harris, Frederick H. and Hayes, J. and Heckman, Timothy M. and Hennessy, G. S. and Hindsley, Robert B. and Holm, Scott and Holmgren, Donald J. and Huang, Chi-hao and Hull, Charles and Husby, Don and Ichikawa, Shin-Ichi and Ichikawa, Takashi and Ivezi{\'c}, {\v Z}eljko and Kent, Stephen and Kim, Rita S. J. and Kinney, E. and Klaene, Mark and Kleinman, A. N. and Kleinman, S. and Knapp, G. R. and Korienek, John and Kron, Richard G. and Kunszt, Peter Z. and Lamb, D. Q. and Lee, B. and Leger, R. French and Limmongkol, Siriluk and Lindenmeyer, Carl and Long, Daniel C. and Loomis, Craig and Loveday, Jon and Lucinio, Rich and Lupton, Robert H. and MacKinnon, Bryan and Mannery, Edward J. and Mantsch, P. M. and Margon, Bruce and McGehee, Peregrine and McKay, Timothy A. and Meiksin, Avery and Merelli, Aronne and Monet, David G. and Munn, Jeffrey A. and Narayanan, Vijay K. and Nash, Thomas and Neilsen, Eric and Neswold, Rich and Newberg, Heidi Jo and Nichol, R. C. and Nicinski, Tom and Nonino, Mario and Okada, Norio and Okamura, Sadanori and Ostriker, Jeremiah P. and Owen, Russell and Pauls, A. George and Peoples, John and Peterson, R. L. and Petravick, Donald and Pier, Jeffrey R. and Pope, Adrian and Pordes, Ruth and Prosapio, Angela and Rechenmacher, Ron and Quinn, Thomas R. and Richards, Gordon T. and Richmond, Michael W. and Rivetta, Claudio H. and Rockosi, Constance M. and Ruthmansdorfer, Kurt and Sandford, Dale and Schlegel, David J. and Schneider, Donald P. and Sekiguchi, Maki and Sergey, Gary and Shimasaku, Kazuhiro and Siegmund, Walter A. and Smee, Stephen and Smith, J. Allyn and Snedden, S. and Stone, R. and Stoughton, Chris and Strauss, Michael A. and Stubbs, Christopher and SubbaRao, Mark and Szalay, Alexander S. and Szapudi, Istvan and Szokoly, Gyula P. and Thakar, Anirudda R. and Tremonti, Christy and Tucker, Douglas L. and Uomoto, Alan and Vanden Berk, Dan and Vogeley, Michael S. and Waddell, Patrick and Wang, Shu-i. and Watanabe, Masaru and Weinberg, David H. and Yanny, Brian and Yasuda, Naoki and {SDSS Collaboration}}, + year = {2000}, + month = sep, + journal = {The Astronomical Journal}, + volume = {120}, + pages = {1579--1587}, + issn = {0004-6256}, + doi = {10.1086/301513}, + urldate = {2023-09-25}, + abstract = {The Sloan Digital Sky Survey (SDSS) will provide the data to support detailed investigations of the distribution of luminous and nonluminous matter in the universe: a photometrically and astrometrically calibrated digital imaging survey of {$\pi$} sr above about Galactic latitude 30\textdegree{} in five broad optical bands to a depth of g'\textasciitilde 23 mag, and a spectroscopic survey of the approximately 106 brightest galaxies and 105 brightest quasars found in the photometric object catalog produced by the imaging survey. This paper summarizes the observational parameters and data products of the SDSS and serves as an introduction to extensive technical on-line documentation.}, + keywords = {/unread,Astrophysics,Cosmology: Observations,Instrumentation: Miscellaneous}, + annotation = {ADS Bibcode: 2000AJ....120.1579Y}, + file = {/Users/nord/Zotero/storage/I2R2RDW3/York et al. - 2000 - The Sloan Digital Sky Survey Technical Summary.pdf} +} + + +@Article{ harris2020NumPy, title = {Array programming with {NumPy}}, - author = {Charles R. Harris and K. Jarrod Millman and St{\'{e}}fan J. - van der Walt and Ralf Gommers and Pauli Virtanen and David - Cournapeau and Eric Wieser and Julian Taylor and Sebastian - Berg and Nathaniel J. Smith and Robert Kern and Matti Picus - and Stephan Hoyer and Marten H. van Kerkwijk and Matthew - Brett and Allan Haldane and Jaime Fern{\'{a}}ndez del - R{\'{i}}o and Mark Wiebe and Pearu Peterson and Pierre - G{\'{e}}rard-Marchant and Kevin Sheppard and Tyler Reddy and - Warren Weckesser and Hameer Abbasi and Christoph Gohlke and - Travis E. Oliphant}, + author = {Charles R. Harris and K. Jarrod Millman and St{\'{e}}fan J. van der Walt and Ralf Gommers and Pauli Virtanen and David Cournapeau and Eric Wieser and Julian Taylor and Sebastian Berg and Nathaniel J. Smith and Robert Kern and Matti Picus and Stephan Hoyer and Marten H. van Kerkwijk and Matthew Brett and Allan Haldane and Jaime Fern{\'{a}}ndez del R{\'{i}}o and Mark Wiebe and Pearu Peterson and Pierre G{\'{e}}rard-Marchant and Kevin Sheppard and Tyler Reddy and Warren Weckesser and Hameer Abbasi and Christoph Gohlke and Travis E. Oliphant}, year = {2020}, month = sep, journal = {Nature}, @@ -198,27 +663,4 @@ @Article{ harris2020array doi = {10.1038/s41586-020-2649-2}, publisher = {Springer Science and Business Media {LLC}}, url = {https://doi.org/10.1038/s41586-020-2649-2} -} - -@article{Ringermacher_2009, - doi = {10.1111/j.1365-2966.2009.14950.x}, - - url = {https://doi.org/10.1111%2Fj.1365-2966.2009.14950.x}, - - year = 2009, - month = {jul}, - - publisher = {Oxford University Press ({OUP})}, - - volume = {397}, - - number = {1}, - - pages = {164--171}, - - author = {Harry I. Ringermacher and Lawrence R. Mead}, - - title = {A new formula describing the scaffold structure of spiral galaxies}, - - journal = {Monthly Notices of the Royal Astronomical Society} -} +} \ No newline at end of file diff --git a/paper/paper.md b/paper/paper.md index fa08c1e..d7b54d2 100644 --- a/paper/paper.md +++ b/paper/paper.md @@ -7,18 +7,26 @@ tags: - simulation - machine learning authors: - - name: Maggie Voetberg - orcid: 0000-0000-0000-0000 + - name: M. Voetberg + orcid: 0009-0005-2715-4709 equal-contrib: true affiliation: "1" - - name: Ashia Lewis - orcid: 0000-0000-0000-0000 + - name: Ashia Livaudais + orcid: 0000-0003-3734-335X equal-contrib: true affiliation: "1" + - name: Becky Nevin + orcid: 0000-0003-1056-8401 + equal-contrib: false + affiliation: "1" + - name: Omari Paul + orcid: 0009-0005-8713-2077 + equal-contrib: false + affiliation: "1" - name: Brian Nord orcid: 0000-0001-6706-8972 equal-contrib: false - affiliation: "1, 2, 3, 4" # (Multiple affiliations must be quoted) + affiliation: "1, 2, 3" # (Multiple affiliations must be quoted) affiliations: - name: Fermi National Accelerator Laboratory, P.O. Box 500, Batavia, IL 60510 index: 1 @@ -26,68 +34,83 @@ affiliations: index: 2 - name: Kavli Institute for Cosmological Physics, University of Chicago, 5801 S Ellis Ave, Chicago, IL 60637 index: 3 - - name: Laboratory for Nuclear Science, MIT, Cambridge MA, 02139-4307 - index: 4 -date: 10 April 2023 +date: 23 June 2023 bibliography: paper.bib --- # Summary -We introduce **DeepBench**, a python library that generates simple simulated image data from first principles, such as basic geometric shapes and astronomical objects. These data are highly valuable for developing (calibration, testing, and benchmarking) statistical and machine learning models because they make it possible to connect the final data product to physically interpretable inputs. This software includes tools to curate and store the datasets to maximize reproducibility. +We introduce **DeepBench**, a Python library that employs mechanistic models (i.e., analytic mathematical models) to simulate data that represent physics-related objects and systems: geometric shapes (e.g., polygon), physics objects (e.g., pendulum), and astronomical objects (e.g., elliptical galaxy). These data take the form of images (two-dimensional) or time series (one-dimensional). In contrast to natural image benchmarks and complex physics simulations, these data have simple, direct, numerical, and traceable connections between the input data and the label data. When seeking a quantifiable interpretation, this kind of data is uniquely suitable for developing, calibrating, testing, and benchmarking statistical and machine learning models. Finally, this software includes methods to curate and store these datasets to maximize reproducibility. # Statement of Need +There are multiple open problems and issues that are critical for the machine learning and scientific communities to address -- principally, interpretability, explainability, uncertainty quantification, and inductive bias in machine learning models when they are applied to scientific data. Multiple kinds of data sets and data simulation software can be used for developing models and confronting these challenges. These data sets range from natural images and text to multi-dimensional data of physical processes. Indeed, multiple benchmark data and simulation software have been created and inculcated for developing and comparing models. However, these benchmarks are typically limited in significant ways. Natural image data sets comprising images from the real or natural world (e.g., vehicles, animals, landscapes) are widely used in the development of machine learning models. These kinds of data sets tend to be large, diverse, and carefully curated. However, they are not underpinned by or constructed upon physical principles: they cannot be generated by mathematical expressions of formal physical theory, so there is not a robust connection between the data and a quantitative theory. Therefore, these data sets have a severely limited capacity to help address many questions in machine learning models, such as uncertainty quantification. On the other hand, complex physics simulations (e.g., cosmological n-body simulations and particle physics simulators) are accurate, detailed, and based on precise quantitative theories and models. This facilitates studies of interpretability and uncertainty quantification because there is the possibility of linking the simulated data to the input choices through each layer of calculation in the simulator. However, they are relatively small in size and number, and they are computationally expensive to reproduce. In addition, while they are underpinned by specific physical functions, the complexity of the calculations makes them challenging as a venue through which to make connections between machine learning results and input choices. Complex physics simulations have one or more layers of mechanistic models. Mechanistic models are defined with analytic functions and equations that describe and express components of a given physical process: these are based on theory and empirical observations. In both of these scenarios, it is difficult to build interpretable models that connect raw data and labels, and it is difficult to generate new data rapidly. -The astronomy community at large is experiencing a lack of benchmark datasets tailored toward the types of computer vision problems specific to astronomy. One of the main motivations of this project was the production of a tool to softly introduce new practitioners of machine learning to simulation datasets, producing datasets that would be easy to integrate with out-of-the-box machine learning frameworks. +The physical sciences community lacks sufficient datasets and software as benchmarks for the development of statistical and machine learning models. In particular, there currently does not exist simulation software that generates data underpinned by physical principles and that satisfies the following criteria: -* replicability, reproducibility -* simple objects -- simpler than most simulators -* physically or geometrically useful objects -* good metadata tracking -* simple data that can be used for benchmarking and for education +* multi-domain +* multi-purpose +* fast +* reproducible +* extensible +* based on mechanistic models +* include detailed noise prescriptions. ## Related Work -Benchmarking and dataset generation is heavily used in the field of Machine Learning. There many benchmark datasets based on natural images -- e.g., MNIST, CIFAR, Imagenet. There are multiple codebases to generate simulated data for astronomy -- e.g., lenstronomy, astropy, galsim, deeplenstronomy. However, these tools for astronomy are complex and difficult to get up and running. The work that is most closely related the work described here is SHAPES `DBLP:journals/corr/WuTWSDH16` for its use of collections of geometric objects as a benchmark with varying levels of complexity. +There are many benchmarks -- datasets and simulation software -- widely used for model building in machine learning, statistics, and the physical sciences. +First, benchmark datasets of natural images include MNIST `[@dengMnistDatabaseHandwritten2012c]`, CIFAR `[@krizhevskyCIFAR10CanadianInstitute2017a]`, Imagenet `[@russakovskyImageNetLargeScale2015a]`. Second, there are several large astronomical observation datasets -- CfA Redshift Survey `[@huchraSurveyGalaxyRedshifts1983]`, Sloan Digital Sky Survey `[@yorkSloanDigitalSky2000]`, and Dark Energy Survey `[@abbottDARKENERGYSURVEY]`. Third, many n-body cosmology simulation data sets serve as benchmarks -- e.g., Millennium `[@springelCosmologicalSimulationCode2005]`, Illustris `[@vogelsbergerIntroducingIllustrisProject2014b]`, EAGLE `[@schayeEAGLEProjectSimulating2015]`, Coyote `[@heitmannCoyoteUniversePrecision2010]`, Bolshoi `[@klypinDARKMATTERHALOS2011]`, CAMELS `[@villaescusa-navarroCAMELSProjectCosmology2021]`, Quijote `[@villaescusa-navarroQuijoteSimulations2020]`. Fourth, there have been multiple astronomy data set challenges that can be considered benchmarks for analysis comparison -- e.g., PLAsTiCC `[@hlozekResultsPhotometricLSST2020a]`, The Great08 Challenge `[@bridleHandbookGREAT08Challenge2009a]`, and the Strong Gravitational Lens Challenge `[@metcalfStrongGravitationalLens2019c]`. Fifth, there are multiple software that generate simulated data for astronomy and cosmology -- e.g., astropy `[@theastropycollaborationAstropyCommunityPython2013a]`, galsim `[@roweGalSimModularGalaxy2015]`, lenstronomy `[@birrerLenstronomyMultipurposeGravitational2018a]`, deeplenstronomy `[@morganDeeplenstronomyDatasetSimulation2021a]`, CAMB `[@CAMBInfo, @lewisEfficientComputationCMB2000]`, Pixell `[@WelcomePixellDocumentation]`, SOXs `[@SOXSSimulatedObservations]`. Finally, particle physics projects use standard codebases for simulations -- e.g., GEANT `[@piaGeant4ScientificLiterature2009]`, GENIE `[@andreopoulosGENIENeutrinoMonte2015]`, and PYTHIA `[@sjostrandPYTHIAEventGenerator2020]`. These simulations span wide ranges in speed, code complexity, and physical fidelity and detail. Unfortunately, these data and software lack a combination of critical features, including mechanistic models, speed, reproducibility, which are needed for more fundamental studies of statistical and machine learning models. The work in this paper is most closely related to SHAPES `[@wuVisualQuestionAnswering2016a]` because that work also uses collections of geometric objects with varying levels of complexity as a benchmark. + + + -# Features +# DeepBench Software -* Simplicity: To avoid needless complexity in execution and use of the program, the program was designed with the philosophy of being easy enough for someone with minimal programming experience to use. -* Reproducibility of Data : Including methods to catalog all produced images and provide a description of all generated objects included in the images. -metadata tracking to help with benchmarking -* Diagnosis: The software is also intended for use as a diagnostic tool, for help in identifying weaknesses during model developmenta. -When working with a difficult problem, being able to incrementally scale back the difficulty (remove noise, decrease the number of classes, increase the class balance, etc.) can be a useful tool in finding the point of difficulty when training a new architecture. The design of **DeepBench** is such that configuration files can be generated with varying levels of complexity, making it possible to “roll back” the scale of a problem, such that possible areas of difficulty can be removed and added back in incrementally, allowing for problems to be solved one by one. +The **DeepBench** software simulates data for analysis tasks that require precise numerical calculations. First, the simulation models are fundamentally mechanistic -- based on relatively simple analytic mathematical expressions, which are physically meaningful. This means that for each model, the number of input parameters that determine a simulation output is small (<$10$ for most models). These elements make the software fast and the outputs interpretable -- conceptually and mathematically relatable to the inputs. Second, **DeepBench** also includes methods to precisely prescribe noise for inputs, which are propagated to outputs. This permits studies and the development of statistical inference models that require uncertainty quantification, which is a significant challenge in modern machine learning research. Third, the software framework includes features that permit a high degree of reproducibility: e.g., random seeds at every key stage of input, a unique identification tag for each simulation run, tracking and storage of metadata (including input parameters) and the related outputs. Fourth, the primary user interface is a YAML configuration file, which allows the user to specify every aspect of the simulation -- e.g., types of objects, numbers of objects, noise type, and number of classes. This feature -- which is especially useful when building and studying complex models like deep learning neural networks -- permits the user to incrementally decrease or increase the complexity of the simulation with a high level of granularity. -# Modules +**DeepBench** has the following features: -![Algorithm Flow overview](figures/overview_diagram.png) +* Exact reproducibility +* Noise and error propagation +* Mechanistic modeling +* Physical sciences-based modeling +* Computational efficiency +* Simulations relevant to multiple domains +* Outputs of varying dimensions +* Readily extensible to new physics and outputs -An overview of the **DeepBench** process. -Dataset parameters, such as the type of objects in each image, and the associated qualities of each object, is passed to the catalog module by the user. -These are collected and used by the Image module to call each individual object with their specified parameters, combined into one composite image, and noise, specified in the image parameters, is then applied before the image is saved. The parameters of each image is stored and saved as well. +# Primary Modules +* Geometry objects: two-dimensional images generated with `matplotlib` `[@hunterMatplotlib2DGraphics2007b]`. The shapes include $N$-sided polygons, arcs, straight lines, and ellipses. They are solid, filled or unfilled two-dimensional shapes with edges of variable thickness. +* Physics objects: one-dimensional profiles for two types of implementations of pendulum dynamics: one using Newtonian physics, the other using Hamiltonian. +* Astronomy objects: two-dimensional images generated based on radial profiles of typical astronomical objects. The star object is created using the Moffat distribution provided by the AstroPy `[@theastropycollaborationAstropyCommunityPython2013a]` library. The spiral galaxy object is created with the function used to produce a logarithmic spiral `[@ringermacherNewFormulaDescribing2009a]`. The elliptical Galaxy object is created using the Sérsic profile provided by the AstroPy library. Two-dimensional models are representations of astronomical objects commonly found in data sets used for galaxy morphology classification. +* Image: two-dimensional images that are combinations and/or concatenations of Geometry or Astronomy objects. The combined images are within `matplotlib` meshgrid objects. Sky images are composed of any combination of Astronomy objects, while geometric images comprise individual geometric shape objects. +* Collection: Provides a framework for producing module images or objects at once and storing all parameters that were included in their generation, including exact noise levels, object hyper-parameters, and non-specified defaults. -* Geometric objects: are generated by utilizing the library matplotlib `Hunter:2007`, and are able to either be left as solid two-dimensional shapes or of an outline with a varying thickness. Shapes include: Rectangles, n sided regular polygons, arcs. straight lines, ellipses, circles. Shapes are limited to two-color rendering, such that all shapes are composed of 0 and 1 values within an array. They can be combined in multiple ways using the "Image" module to produce composite images with both different geometric objects and astronomical objects. -* Physics objects: The N-Body object is currently the sole one-dimensional model available in the alpha release. The N-Body model is output as a set of NumPy `harris2020array` arrays containing coordinates representing the path of the points produced, along with the kinetic and potential energy produced over the specified duration. -* Astronomical Objects: Astronomical objects offer simplified renderings of common profiles found in astronomical data sets. Two-dimensional models are representations of astronomical objects commonly found in data sets used for galaxy morphology classification. All objects also come with the option to append various levels of Gaussian and Poisson noise, and are output as NumPy arrays. The Star object is created using the Moffat distribution provided by the AstroPy `astropy:2018` modeling library. The Elliptical Galaxy object is created using the Sérsic profile provided by the AstroPy modelling library. The profile of the Spiral Galaxy object is created by simulating the function used to produce a logarithmic spiral `Ringermacher_2009` -* Image The image module allows users to concatenate various shape or astronomical objects within a matplotlib meshgrid object in order to simulate the profiles and shape distributions commonly seen in images used in more complex astronomical data sets. Three distinct image types are available - sky images, lensing images, and geometric images. Sky images are composed of any combination of user-specified galaxy and star objects, while lensing images are randomized combinations of arc and star objects. Lastly, geometric images are canvased assortments of any of the individual geometric shape objects available. -* Catalogue: The Catalogue module allows users to specify the size, contents, and output directory of a data set composed of **DeepBench**'s available image types or individual object images. Catalogues can either be entirely randomized, with parameters of the included image being randomly chosen, or have parameters specified by the user at various levels of granularity. The only required argument in the creation of Catalogue objects are the output directory and image type. Within the Catalogue module also exists a Collection class, which concatenates various catalogues of varying image types into one data set. Catalogues and collections can be output as a directory of.jpeg files, .npy, or .h5. +All objects also come with the option to add noise to each object. For Physics objects -- i.e., the pendulum -- the user may add Gaussian noise to parameters: initial angle $\theta_0$, the pendulum length $L$, the gravitational acceleration $g$, the planet properties $\Phi = (M/r^2)$, and Newton's gravity constant $G$. Note that $g = G * \Phi = G * M/r^2$: all parameters in that relationship can receive noise. For Astronomy and Geometry Objects, which are images, the user can add Poisson or Gaussian noise to the output images. Finally, the user can regenerate the same noise using the saved random seed. +# Example Outputs + +![Example outputs of **DeepBench**, containing shapes, astronomy objects, and the two pendulum implementations. Variants include a single object, a noisy single object, two objects, and two noisy objects. Pendulums show noisy and non-noisy variants of the Newtonian (left) and Hamiltonian (right) mathematical simulations.](figures/example_objects.png) + +![Example physics simulations from **DeepBench**. Pendulums show noisy and non-noisy variants of the Newtonian (left) and Hamiltonian (right) mathematical simulations.](figures/pendulums.png) # Acknowledgements -We acknowledge contributions from Alex Ciprijanovic, Renee Hlozek, Brechmos. +*M. Voetberg*: conceptualization, methodology, software, writing, project administration. *Ashia Livaudais*: conceptualization, methodology, software, writing, project administration. *Becky Nevin*: software, project administration. *Omari Paul*: software. *Brian Nord*: conceptualization, methodology, project administration, funding acquisition, supervision, writing. + +We acknowledge contributions from Alex \'Ciprijanovi\'c, Renee Hlozek, Craig Brechmos. Work supported by the Fermi National Accelerator Laboratory, managed and operated by Fermi Research Alliance, LLC under Contract No. DE-AC02-07CH11359 with the U.S. Department of Energy. The U.S. Government retains and the publisher, by accepting the article for publication, acknowledges that the U.S. Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for U.S. Government purposes. -We acknowledge the Deep Skies Lab as a community of multi-domain experts and collaborators who’ve facilitated an environment of open discussion, idea-generation, and collaboration. This community was important for the development of this project. +We acknowledge the Deep Skies Lab as a community of multi-domain experts and collaborators who have facilitated an environment of open discussion, idea-generation, and collaboration. This community was important for the development of this project. # References + diff --git a/pyproject.toml b/pyproject.toml index 82c61ac..bfcf53e 100644 --- a/pyproject.toml +++ b/pyproject.toml @@ -4,11 +4,11 @@ version = "0.2.2" description = "Physics Benchmark Dataset Generator" license = "Apache 2.0" authors = ["M. Voetberg ", "Ashia Livaudais", "Becky Nevin", "Omari Paul", "Brian Nord"] -maintainers = ["M. Voetberg "] +maintainers = ["M. Voetberg ", "Brian Nord =1.0.0"]