Publications

This paper describes the implementation of semi-structured deep distributional regression, a flexible framework to learn distributions based on a combination of additive regression models and deep neural networks. deepregression is implemented in both R and Python, using the deep learning libraries TensorFlow and PyTorch, respectively. The implementation consists of (1) a modular neural network building system for the combination of various statistical and deep learning approaches, (2) an orthogonalization cell to allow for an interpretable combination of different subnetworks as well as (3) pre-processing steps necessary to initialize such models. The software package allows to define models in a user-friendly manner using distribution definitions via a formula environment that is inspired by classical statistical model frameworks such as mgcv. The packages' modular design and functionality provides a unique resource for rapid and reproducible prototyping of complex statistical and deep learning models while simultaneously retaining the indispensable interpretability of classical statistical models.

In an early study of the properties and capabilities of the multiregion, relaxed magnetohydrodynamic model, Hole, Hudson & Dewar claim that they are able to construct a multiregion stepped pressure cylindrical equilibrium which does not require the existence of surface currents. We present a brief argument showing that this claim is incorrect, and clarify the meaning of their statement. Furthermore, even with the statement clarified, we demonstrate that it is not possible to find solutions to reproduce the equilibrium corresponding to the parameters given in the article. We invite the authors to provide a corrigendum with the correct values of the equilibrium they constructed.

Each year vast international resources are wasted on irreproducible research. The scientific community has been slow to adopt standard software engineering practices, despite the increases in high-dimensional data, complexities of workflows, and computational environments. Here we show how scientific software applications can be created in a reproducible manner when simple design goals for reproducibility are met. We describe the implementation of a test server framework and 40 scientific benchmarks, covering numerous applications in Rosetta bio-macromolecular modeling. High performance computing cluster integration allows these benchmarks to run continuously and automatically. Detailed protocol captures are useful for developers and users of Rosetta and other macromolecular modeling tools. The framework and design concepts presented here are valuable for developers and users of any type of scientific software and for the scientific community to create reproducible methods. Specific examples highlight the utility of this framework and the comprehensive documentation illustrates the ease of adding new tests in a matter of hours.

The complex interplay of magnetohydrodynamics, gravity, and supersonic turbulence in the interstellar medium (ISM) introduces non-Gaussian structure that can complicate comparison between theory and observation. We show that the Wavelet Scattering Transform (WST), in combination with linear discriminant analysis (LDA), is sensitive to non-Gaussian structure in 2D ISM dust maps. WST-LDA classifies magnetohydrodynamic (MHD) turbulence simulations with up to a 97\% true positive rate in our testbed of 8 simulations with varying sonic and Alfvénic Mach numbers. We present a side-by-side comparison with two other methods for non-Gaussian characterization, the Reduced Wavelet Scattering Transform (RWST) and the 3-Point Correlation Function (3PCF). We also demonstrate the 3D-WST-LDA and apply it to classification of density fields in position-position-velocity (PPV) space, where density correlations can be studied using velocity coherence as a proxy. WST-LDA is robust to common observational artifacts, such as striping and missing data, while also sensitive enough to extract the net magnetic field direction for sub-Alfvénic turbulent density fields. We include a brief analysis of the effect of point spread functions and image pixelization on 2D-WST-LDA applied to density fields, which informs the future goal of applying WST-LDA to 2D or 3D all-sky dust maps to extract hydrodynamic parameters of interest.

Non-negative Matrix Factorization (NMF) methods offer an appealing unsupervised learning method for real-time analysis of streaming spectral data in time-sensitive data collection, such as in situ characterization of materials. However, canonical NMF methods are optimized to reconstruct a full dataset as closely as possible, with no underlying requirement that the reconstruction produces components or weights representative of the true physical processes. In this work, we demonstrate how constraining NMF weights or components, provided as known or assumed priors, can provide significant improvement in revealing true underlying phenomena. We present a PyTorch based method for efficiently applying constrained NMF and demonstrate this on several synthetic examples. When applied to streaming experimentally measured spectral data, an expert researcher-in-the-loop can provide and dynamically adjust the constraints. This set of interactive priors to the NMF model can, for example, contain known or identified independent components, as well as functional expectations about the mixing of components. We demonstrate this application on measured X-ray diffraction and pair distribution function data from in situ beamline experiments. Details of the method are described, and general guidance provided to employ constrained NMF in extraction of critical information and insights during in situ and high-throughput experiments.

We present an efficient implementation of periodic Gaussian density fitting (GDF) using the Coulomb metric. The three-center integrals are divided into two parts by range-separating the Coulomb kernel, with the short-range part evaluated in real space and the long-range part in reciprocal space. With a few algorithmic optimizations, we show that this new method -- which we call range-separated GDF (RSGDF) -- scales sublinearly to linearly with the number of k-points for small to medium-sized k-point meshes that are commonly used in periodic calculations with electron correlation. Numerical results on a few three-dimensional solids show about 10-fold speedups over the previously developed GDF with little precision loss. The error introduced by RSGDF is about 10

The performance of objective image quality assessment (IQA) models has been evaluated primarily by comparing model predictions to human quality judgments. Perceptual datasets gathered for this purpose have provided useful benchmarks for improving IQA methods, but their heavy use creates a risk of overfitting. Here, we perform a large-scale comparison of IQA models in terms of their use as objectives for the optimization of image processing algorithms. Specifically, we use eleven full-reference IQA models to train deep neural networks for four low-level vision tasks: denoising, deblurring, super-resolution, and compression. Subjective testing on the optimized images allows us to rank the competing models in terms of their perceptual performance, elucidate their relative advantages and disadvantages in these tasks, and propose a set of desirable properties for incorporation into future IQA models.

We describe a new algorithm to solve the time dependent, frequency integrated radiation transport (RT) equation implicitly, which is coupled to an explicit solver for equations of magnetohydrodynamics (MHD) using {\sf Athena++}. The radiation filed is represented by specific intensities along discrete rays, which are evolved using a conservative finite volume approach for both cartesian and curvilinear coordinate systems. All the terms for spatial transport of photons and interactions between gas and radiation are calculated implicitly together. An efficient Jacobi-like iteration scheme is used to solve the implicit equations. This removes any time step constrain due to the speed of light in RT. We evolve the specific intensities in the lab frame to simplify the transport step. The lab-frame specific intensities are transformed to the co-moving frame via Lorentz transformation when the source term is calculated. Therefore, the scheme does not need any expansion in terms of v/c. The radiation energy and momentum source terms for the gas are calculated via direct quadrature in the angular space. The time step for the whole scheme is determined by the normal Courant -- Friedrichs -- Lewy condition in the MHD module. We provide a variety of test problems for this algorithm including both optically thick and thin regimes, and for both gas and radiation pressure dominated flows to demonstrate its accuracy and efficiency.

Active galactic nuclei (AGNs) are powered by geometrically thin accretion disks surrounding a central supermassive black hole. Here we explore the evolution of stars embedded in these extreme astrophysical environments (AGN stars). Because AGN disks are much hotter and denser than most components of the interstellar medium, AGN stars are subject to very different boundary conditions than normal stars. They are also strongly affected by both mass accretion, which can run away given the vast mass of the disk, and mass loss due to super-Eddington winds. Moreover, chemical mixing plays a critical role in the evolution of these stars by allowing fresh hydrogen accreted from the disk to mix into their cores. We find that, depending on the local AGN density and sound speed and the duration of the AGN phase, AGN stars can rapidly become very massive (M > 100 M⊙). These stars undergo core collapse, leave behind compact remnants, and contribute to polluting the disk with heavy elements. We show that the evolution of AGN stars can have a profound impact on the evolution of AGN metallicities, as well as the production of gravitational wave sources observed by LIGO-Virgo. We point to our Galactic Center as a region well suited to testing some of our predictions for this exotic stellar evolutionary channel.

Latent Gaussian copula models provide a powerful means to perform multi-view data integration since these models can seamlessly express dependencies between mixed variable types (binary, continuous, zero-inflated) via latent Gaussian correlations. The estimation of these latent correlations, however, comes at considerable computational cost, having prevented the routine use of these models on high-dimensional data. Here, we propose a new computational approach for estimating latent correlations via a hybrid multi-linear interpolation and optimization scheme. Our approach speeds up the current state of the art computation by several orders of magnitude, thus allowing fast computation of latent Gaussian copula models even when the number of variables p is large. We provide theoretical guarantees for the approximation error of our numerical scheme and support its excellent performance on simulated and real-world data. We illustrate the practical advantages of our method on high-dimensional sparse quantitative and relative abundance microbiome data as well as multi-view data from The Cancer Genome Atlas Project. Our method is implemented in the R package mixedCCA, available at https://github.com/irinagain/mixedCCA this https URL.