AccretR calculates mass, radius, and bulk composition along a specified growth track for orderly/hierarchical, runaway, and random particle accretion models. Elements in the model include concentrations of H, C, N, O, Na, Mg, Al, Si, S, Cl, K, Ca, and Fe. Maximal water is also computed, assuming all H goes into forming water. Accretional heat is also calculated. AccretR is optimized to build Jupiter's moon Europa, and Saturn's moons Titan and Enceladus, from CI, CM, CR, CK, CO and CV carbonaceous chondrite meteorites, cometary material (using comet 67P/Churyumov-Gerasimenko), and pure water ice.

THAI analyzes and visualizes climate model output for the TRAPPIST-1 Habitable Atmosphere Intercomparison (THAI) project, which examines TRAPPIST-1e under several different atmosphere scenarios. The package includes functions to preprocess and clean the data and common and model-specific variables for convenience. THAI processes and plots the data, allowing for examination and intercomparison of results from the different models.

Pigi (Parallel Interferometric GPU Imager) implements the image domain gridding algorithm and is compatible with both NVIDIA and AMD graphics cards. It provides a high-performance implementation capable of gridding hundreds of mega visibilities per second on modest hardware. The code can correct for baseline-, time-, and direction-dependent effects such as the primary beam or ionosphere as part of the (de)gridding process. Pigi provides end-to-end deconvolution capabilities with a basic iterative cleaning implementation.

legacypipe produces DESI Legacy Imaging Surveys (aka the Legacy Surveys). It can process individual exposures from many cameras, including the Dark Energy Camera on the Blanco telescope, the 90Prime camera on the Bok telescope, and the Mosaic3 camera on the Mayall telescope. The code can also process exposures from the Hyper-SuprimeCam on Subaru, the old SuprimeCam on Subaru, MegaCam on the Canada-France-Hawaii Telescope, and image products from the GALEX and WISE satellites. Legacypipe performs source detection, and then measurement via forward-modeling using The Tractor (ascl:1604.008). It generates coadded output images as well as catalogs, plus a variety of metrics useful for understanding the properties of the imaging.

The Python code smhr (Spectroscopy Made Harder) wraps the MOOG spectral synthesis code (ascl:1202.009) to analyze high-resolution stellar spectra. It offers numerous analysis tools, including normalization of apertures, inverse variance-weighted stitching of overlapping apertures and/or sequential exposures. The code also provides Doppler measurement and correction, automatic measurement of EWs, and multiple methods for inferring stellar parameters; further, it measures elemental abundances from EWs or spectral synthesis and performs a rigorous uncertainty analysis. smhr can be run automatically (in batch mode) or interactively through a graphical user interface. Analyses can be saved to a single file for, for example, distribution to other spectroscopists or release with a publication.

Based on oxkat (ascl:2009.003), polkat focuses on automating full polarization calibration and snapshot (i.e., second-scale) imaging of polarimetric radio data taken with the MeerKAT telescope. Accepting raw visibilities in Measurement Set format, polkat performs the necessary data editing, calibration (reference and self-calibration), and imaging to extract the complete polarization properties for user-defined target sources. Required software packages, including, but not limited to, CASA (ascl:1107.013), WSClean (ascl:1408.023), and QuartiCal (ascl:2305.006) are containerized with Apptainer/Singularity. polkat can be run locally or on high-performance computing that uses a slurm job scheduler; for the latter option, polkat will generate the necessary job submission files.

Sledgehamr (ScaLar fiEld Dynamics Getting solvEd witH Adaptive Mesh Refinement) simulates the dynamics of coupled scalar fields on a 3-dimensional mesh. Adaptive mesh refinement (AMR) can boost performance if spatially localized regions of the scalar field require high resolution. sledgehamr is compatible with both GPU and CPU clusters, and, because it is AMReX-based (ascl:2409.012), offers a flexible and customizable framework. This framework enables various applications, such as the generation of gravitational wave spectra.

cosmocnc evaluates the number count likelihood of galaxy cluster catalogs. Fast Fourier Transform (FFT) convolutions are used to evaluate some of the likelihood integrals. The code supports three types of likelihoods (unbinned, binned, and an extreme value likelihood); it also supports the addition of stacked cluster data (e.g., stacked lensing profiles), which is modeled in a consistent way with the cluster catalog. The package produce mass estimates for each cluster in the sample, which are derived assuming the hierarchical model that is used to model the mass observables, and generates synthetic cluster catalogs for a given observational set-up. cosmocnc interfaces with the Markov chain Monte Carlo (MCMC) code Cobaya (ascl:1910.019), allowing for easy-to-run MCMC parameter estimation.

SZiFi (pronounced "sci-fi") implements the iterative multi-frequency matched filter (iMMF) galaxy cluster finding method. It can be used to detect galaxy clusters with mm intensity maps through their thermal Sunyaev-Zeldovich (tSZ) signal. As a novel feature, SZiFi can perform foreground deprojection via a spectrally constrained MMF or sciMMF, and can also be used for point source detection.

hmvec is a pure Python/numpy vectorized general halo model and HOD code. It includes support for 3d power spectra involving NFW, Battaglia electron density profiles and galaxy HODs. It also supports 2d power spectra including tSZ, cosmic shear, galaxy-galaxy lensing and CMB lensing. hmvec calculates a vectorized FFT for a given profile over all points in mass and redshift, using one double loop over mass and redshift to interpolate the profile Fourier transforms to the target wavenumbers; every other part of the code is vectorized.

Spectool is a toolkit designed for processing astronomical spectral data, offering a collection of common spectral analysis algorithms. The package includes functions for spectral resampling, spectral flattening, radial velocity measurements, spectral convolution broadening, and more. Each function in the package is implemented independently, allowing users to select and utilize the desired features as needed. The functions are designed with simple and intuitive interfaces, ensuring ease of use for various data sets and analysis tasks.

Hierarchical Semi-Sparse Cube (HiSS-Cube) framework provides highly parallel processing of combined multi-modal multi-dimensional big data. The package builds a database on top of the HDF5 framework which supports parallel queries. A database index on top of HDF5 can be easily constructed in parallel, and the code supports efficient multi-modal big data combinations. The performance of HiSS-Cube is bounded by the I/O bandwidth and I/O operations per second of the underlying parallel file system; it scales linearly with the number of I/O nodes and can be extended to any kind of multidimensional data combination and information retrieval.

NbodyGradient computes gradients of N-body integrations for Newtonian gravity and arbitrary N-body hierarchies. Developed for transit-timing analyses and written in Julia, NbodyGradient gives derivatives of the transit times with respect to the initial conditions, either masses and Cartesian coordinates/velocities or orbital elements.

The spectools_ir suite analyzes medium/high-resolution IR molecular astronomical spectra. It has three main sub-modules (flux_calculator, slabspec, and slab_fitter) and also offers a sub-module (utils) with a few additional functions. Written with infrared medium/high-resolution molecular spectroscopy in mind, spectools_ir generally assumes spectra are in units of Jy and microns and uses information from the HITRAN molecular database. Some routines are more general, but users interested in other applications should proceed with caution.

ROCKE-3D (Resolving Orbital and Climate Keys of Earth and Extraterrestrial Environments with Dynamics) models the atmospheres and oceans of solar system and exoplanetary terrestrial planets. Written in Fortran, it is a three-dimensional General Circulation Model (GCM). ROCKE-3D requires Panoply, the SOCRATES radiation code and spectral files, and has several additional dependencies.

Deep-Transit detects transits using a deep learning based 2D object detection algorithm. The code determines the light curve and outputs the transiting candidates' bounding boxes and confidence scores. It has been trained for Kepler and TESS data, and can be extended to other photometric surveys and even ground-based observations. Deep-Transit also provides an interface for training new datasets.

StellarSpecModel is a Python package to interpolate the stellar spectral grid. Users provide stellar parameters (Teff, FeH, logg), the package will return the corresponding stellar spectrum.

This packagge also designed for generating and analyzing theoretical stellar spectral energy distributions (SEDs). The package includes functionality for both single and binary star systems, incorporating extinction models and the ability to handle photometric data in various filter bands.

Spinifex is a pure Python tooling for ionospheric corrections in radio astronomy, e.g. getting total electron content and rotation measures.

The Galaxy Morphology Posterior Estimation Network (GaMPEN) is a Bayesian machine learning framework that can estimate robust posteriors (i.e., values + uncertainties) for structural parameters of galaxies. GaMPEN also automatically crops input images to an optimal size before structural parameter estimation.

GaMPEN’s predicted posteriors are extremely well-calibrated (less than 5% deviation) and have been shown to be up to 60% more accurate compared to the uncertainties predicted by many light-profile fitting algorithms.

Once trained, it takes GaMPEN less than a millisecond to perform a single model evaluation on a CPU. Thus, GaMPEN’s posterior prediction capabilities are ready for large galaxy samples expected from upcoming large imaging surveys, such as Rubin-LSST, Euclid, and NGRST.

GaMorNet is a Convolutional Neural Network to classify galaxies morphologically. GaMorNet does not need a large amount of training data (as it is trained on simulations and then transfer-learned on a small portion of real data) and can be applied on multiple datasets. Till now, GaMorNet has been tested on ~100,000 SDSS g-band galaxies and ~20,000 CANDELS H-band galaxies and has a misclassification rate of less than 5%

This repository implements an optimized XGBoost-based framework for photometric classification of Type Ia supernovae, addressing class imbalance through PR-AUC and F1-score prioritization. The approach is designed for scalability in large-scale astronomical surveys such as LSST and ensures improved classification robustness compared to traditional metrics like ROC-AUC.

We present a new algorithm for identifying superbubbles in HI column density maps of both observed and simulated galaxies that has only two adjustable parameters. The algorithm includes an automated galaxy-background separation step to focus the analysis on the galactic disk. To test the algorithm, we compare the superbubbles it finds in a simulated galactic disk with the ones it finds in 21~cm observations of a similar galactic disk. The sizes and radial distribution of those superbubbles are indeed qualitatively similar. However, superbubbles in the simulated galactic disk have lower central H~I column densities. The H~I superbubbles in the simulated disk are spatially associated with pockets of hot gas. We conclude that the algorithm is a promising method for systematically identifying and characterizing superbubbles using only HI column density maps that will enable standardized tests of stellar feedback models used in galaxy simulations.

exoscene simulates direct images of exoplanetary systems. Written in Python, the package has three modules. These modules can determine a planet's relative astrometry ephemeris, its phase function, and flux ratio, compute the band-integrated irradiance of a star, and accurately resample an image model array to a detector array. exoscene also offers modeling and mapping functions and has additional capabilities.

AESTRA (Auto-Encoding STellar Radial-velocity and Activity) uses deep learning for precise radial velocity measurements in the presence of stellar activity noise. The architecture combines a convolutional radial-velocity estimator and a spectrum auto-encoder. The input consists of a collection of hundreds or more of spectra of a single star, which span a variety of activity states and orbital motion phases of any potential planets. AESTRA does not require any prior knowledge about the star.

APPLESOSS (A Producer of ProfiLEs for SOSS) builds 2D spatial profiles for the first, second, and third diffraction orders for a NIRISS/SOSS GR700XD/CLEAR observation. The profiles are entirely data driven, retain a high level of fidelity to the original observations, and can be used as the specprofile reference file for ATOCA (ascl:2502.016). They can also be used as a PSF weighting for optimal extractions.

ATMOSPHERIX reads t.fits files from the Canada-France-Hawaii Telescope's near-infrared spectropolarimeter SPIRou, processes the data to remove telluric/stella contributions, and performs the correlation analysis for a given planet atmosphere template. The correlation function computes the correlation between the data and model for a grid of planet velocimetric semi-amplitude and systemic velocity. ATMOSPHERIX takes transmission spectroscopy into account and allows the user to inject a synthetic planet if desired.

Bioverse assesses the diagnostic power of a statistical exoplanet survey of the properties of nearby terrestrial exoplanets via direct imaging or transit spectroscopy. It combines Gaia-based stellar samples with Kepler-derived exoplanet demographics and a mission simulator that enables exploration of a variety of observing, follow-up, and characterization strategies. The code contains a versatile module for population-level hypothesis testing supporting trade studies and survey optimization. Bioverse supports direct imaging or transit missions, and its modularity makes it adaptable to any mission concept that makes measurements on a sample of exoplanets.

The Find Emission LINEs tool FELINE combines a fully parallelized galaxy line template
matching with a matched filter approach for individual emission features.
The FELINE algorithm evaluates the likelihood in each spectrum of a 3D data cube for emission lines at the positions provided by a given redshift and a certain combination of typical emission features.
FELINE does not evaluate the observed data cube directly, but instead utilizes the result of an emission line matched filter to boost the signal-to-noise of any such feature in the data cube. FELINE, however, does not pick individual peaks from that data as separate objects but instead simultaneously assesses the accumulative signal at all spectral positions that correspond to a certain set of emission lines at any redshift within the range of interest.

CROCODILE (CROss-COrrelation retrievals of Directly-Imaged self-Luminous Exoplanets) runs atmospheric retrievals of directly observed gas giant exoplanets by adopting adequate likelihood functions. The code makes use of petitRADTRANS (ascl:2207.014) and PyMultiNest (ascl:1606.005) and provides a statistical framework to interpret the photometry, low-resolution spectroscopy, and medium (and higher) resolution cross-correlation spectroscopy.

DIA (Delta function Difference Imaging Code) provides a difference image analysis pipeline that employs a delta-function kernel; this is useful for reducing TESS Full Frame Images. DIA's scripts are available in both Python and IDL and are nearly identical in their outputs. Together, the scripts make a pipeline that cleans and aligns images, generates a master frame by combining all available images, performs image subtraction, generates light curves, and does a basic detrending to the light curves based on magnitude. DIA can also apply bias subtraction, flat fielding, background subtraction and align images to the first image in the list.

Given mass, radius, and equilibrium temperature, ExoMDN can deliver a full posterior distribution of mass fractions and thicknesses of each planetary layer. A machine-learning model for the interior characterization of exoplanets based on Mixture Density Networks (MDN), ExoMDN is trained on a large dataset of more than 5.6 million synthetic planets below 25 Earth masses. These synthetic planets consist of an iron core, a silicate mantle, a water and high-pressure ice layer, and a H/He atmosphere. ExoMDN uses log-ratio transformations to convert the interior structure data into a form that the MDN can easily handle.

GEOCLIM.jl, written in Julia, replicates some features of the original GEOCLIM model written in Fortran. It also extends the original weathering equations WHAK and MAC, which ignore direct dependence on pCO2 and include direct pCO2 dependence respectively. The code estimates global silicate weathering rates from gridded climatology. GEOCLIM.jl estimates weathering during periods of Earth history when the continental configuration was radically different, typically more than 100 million years ago, and includes functions to compute, for example, land/ocean fraction, area-weighted average, area-weighted sum, and land mass perimeter, among other values.

Gollum performs spectral visualization and analysis. It offers both a programmatic interface and a visual interface that help users analyze stellar and substellar spectra, with support included for a set of precomputed synthetic spectral model grids.