CosmosCanvas creates perception-based color maps for different astrophysical properties such as spectral index and velocity fields. Three tutorials demonstrate how to use python code to exploit and adjust the boundaries in these divergent colour schemes. Intended to work with human physiology, each tutorial offers at least one default scheme that is monotonic in value both as a redundancy for supporting data information and an aid for colour blind viewers. This library relies on Gilles Ferrand's colourspace library.

LoSoTo (LOFAR Solution Tool) performs a variety of operations on H5parm data, which is based on the HDF5 format; it isolates direction independent systematic effects and can therefore be transferred to the target field. Subsets of data can be selected for each operation using lists of axes values, regular expressions, or intervals. The LoSoTo package stores solutions in arrays organized in a hierarchical fashion; this provides flexibility and preserves performance. The code can, for example, extract Faraday rotation from RR/LL phase solutions or a rotation matrix, clip solutions around the median, and calculate the ionospheric structure function. LoSoTo includes an outlier flagging procedure, normalizes solutions to a given value, and offers an advanced plotting routine, and many other operations.

deal.II computes solutions to partial differential equations (PDEs) using adaptive finite elements. The code provides an interface for processing PDEs accessible to both laptops and supercomputers, and has been used to investigate the local and global waveform effects of gravitational waves by numerical simulation. deal.II supports massively parallel computing of very large linear systems of equations and provides access to triangulation of various geometries of the simulation domain.

DARC (Dirac Atomic R-matrix Codes) enables the study of continuum processes for a general atomic system. The suite of programs calculate electron-atom or electron-ion collision cross-sections. In addition, the programs include code for bound-state and photoionization calculations.

harmonic learns an approximate harmonic mean estimator (referred to as a "learnt harmonic mean estimator") from posterior distribution samples to compute the marginal likelihood required for Bayesian model selection. Using a large number of independent Markov chain Monte Carlo (MCMC) chains from another package such as emcee (ascl:1303.002), harmonic uses importance sampling to learn a new target distribution in order to optimize an approximate harmonic estimator while minimizing its variance.

The Feed Forward Neural Network SYSNet models the relationship between the imaging maps, such as stellar density and the observed galaxy density field, in order to mitigate the systematic effects and to make a robust galaxy clustering measurements. The cost function is Mean Squared Error and a L2 regularization term, and the optimization algorithm is Adaptive Moment (ADAM).

Ostrich emulates surrogate models for complex and expensive functions using Principal Component Analysis (PCA) to decompose a signal, then interpolate the PCA weights over the parameters θ using a Gaussian Process interpolator. The code is trained on samples from the expensive functions, recreating and interpolating between those training samples with reduced computational cost, and recalculating for each use.

baryon-sweep produces a robust outlier rejection while simultaneously preserving the signal of the science target. The code works as a standalone solution or as a supplement to the current pipeline software. baryon-sweep creates the 2D pixel mask and mask layers, processes the sky (non-science target) spaxels, and creates a post-processed cube ready for use.

SolarKAT mitigates solar interference in MeerKAT data and recovers the visibilities rather than discarding them; this solar imaging pipeline takes 1GC calibrated data in Measurement Set format as input. Written in Python, the pipeline employs solar tracking, subtraction, and peeling techniques to enhance data quality by significantly reducing solar radio interference. This is achieved while preserving the flux measurements in the main field. SolarKAT is versatile and can be applied to general radio astronomy observations and solar radio astronomy; additionally, generated solar images can be used for weather forecasting. SolarKAT is deployed in Stimela (ascl:2305.007). It is based on existing radio astronomy software, including CASA (ascl:1107.013), breizorro (ascl:2305.009), WSclean (ascl:1408.023), Quartical (ascl:2305.006), and Astropy (ascl:1304.002).

LoRD (Locate Reconnection Distribution) identifies the locations and structures of 3D magnetic reconnection within discrete magnetic field data. The toolkit contains three main functions; the first, ARD (Analyze Reconnection Distribution) locates the grids undergoing reconnection without null points and also recognizes the local configurations of reconnection sites. ANP (Analyze Null Points) locates and classifies the 3D null points, and APNP (Analyze Projected Null Points) analyzes the 2D neutral points projected on a plane near a cell. LoRD is written in Matlab and the toolkit contains demo scripts.

maskfill inward extrapolates edge pixels just outside masked regions, using iterative median filtering and the full information contained in the edge pixels. This provides seamless transitions between masked pixels and good pixels, and allows high fidelity reconstruction of gaps in continuous narrow features. An image and a mask the only required inputs.

CRR (Convex Ridge Regularizer) takes the gradient of regularizers that are the sum of convex-ridge functions and parameterizes them using a neural network that has a single hidden layer with increasing and learnable activation functions. The neural network is trained within a few minutes as a multistep Gaussian denoiser, and offers improvements for denoising and image reconstruction over other methods with similar reliability.

QuantifAI reconstructs radio interferometric images using scalable Bayesian uncertainty quantification relying on data-driven (learned) priors. It relies on the convex accelerated optimization algorithms in CRR (ascl:2401.016) and is built on top of PyTorch. QuantifAI also includes MCMC algorithms for posterior sampling.

tidalspin uses a Bayesian approach to infer posterior distributions of a black hole's parameters (mass and spin) in an observed tidal disruption event, given a prior estimate of the black hole’s mass (e.g., from a galactic scaling relationship, or the tidal disruption event’s observed properties). These posterior distributions will only utilize the properties of tidal forces in their inference. tidalspin can be applied to the population of tidal disruption events already discovered.

StructureFunction determines the X-ray Structure Function of a population of Active Galactic Nuclei (AGN) for which two epoch X-ray observations are available and are separated by rest frame time interval. The calculation of the X-ray structure function is Bayesian. The sampling of the likelihood uses Stan (ascl:1801.003) for statistical modeling and high-performance statistical computation.

escatter.py performs Monte Carlo simulations of electron scattering events. The code was developed to better understand the emission lines from the interacting supernova SN 2021adxl, specifically the blue excess seen in the Hα 6563A emission line. escatter follows a photon that was formed in a thin interface between the supernova ejecta and surrounding material as it travels radially outwards through the dense material, scattering electrons outwards until it reaches an optically thin region, and plots a histogram of the emergent photons.

NMMA probes nuclear physics and cosmology with multimessenger analysis. This fully featured, Bayesian multi-messenger pipeline targets joint analyses of gravitational-wave and electromagnetic data (focusing on the optical). Using bilby (ascl:1901.011) as the back-end, the software uses a variety of samplers to sampling these data sets. NMMA uses chiral effective field theory based neutron star equation of states when performing inference, and is also capable of estimating the Hubble Constant.

Rfits reads and writes FITS images, tables, and headers. Written in R, Rfits works with all types of compressed images, and both ASCII and binary tables. It uses CFITSIO (ascl:1010.001) for all low level FITS IO, so in general should be as fast as other CFITSIO-based software. For images, Rfits offers fully featured memory mapping and on-the-fly subsetting (by pixel and coordinate) and sparse pixel sampling, allowing for efficient inspection of very large (larger than memory) images.

Rwcs offers access to all the projection and distortion systems of WCSLIB (ascl:1108.003) in R. This can be used directly for, for example, pixel lookups, or for higher level general distortion and projection.

CCBH-Numerics (previously called CCBH-PLPP) computes the probability of the existence of a single cosmologically coupled black hole (BH) with a formation mass below a specified threshold for given observational data of binary black holes (BBHs) from gravitational waves. The code uses the unbiased population of BBHs, as given by the power-law-plus-peak (PLPP) profile, as the observational input, and assumes that the detected BBHs are formed from stellar evolution, not primordial BHs. CCBH-Numerics also works with individual data from BBHs and for NSBH pairs as well.

MGPT (Modified Gravity Perturbation Theory) computes 2-point statistics for LCDM model, DGP and Hu-Sawicky f(R) gravity. Written in C, the code can be easily modified to include other models. Specifically, it computes the SPT matter power spectrum, SPT Lagrangian-biased tracers power spectrum, and the CLPT matter correlation function. MGPT also computes the CLPT Lagrangian-biased tracers correlation function and a set of Q and R functionsfrom which other statistics, as leading order bispectrum, can be constructed.

We introduce GalMOSS, a Python-based, Torch-powered tool for two-dimensional fitting of galaxy profiles. By seamlessly enabling GPU parallelization, GalMOSS meets the high computational demands of large-scale galaxy surveys, placing galaxy profile fitting in the LSST-era. It incorporates widely used profiles such as the Sérsic, Exponential disk, Ferrer, King, Gaussian, and Moffat profiles, and allows for the easy integration of more complex models. Tested on 8,289 galaxies from the Sloan Digital Sky Survey (SDSS) g-band with a single NVIDIA A100 GPU, GalMOSS completed classical Sérsic profile fitting in about 10 minutes. Benchmark tests show that GalMOSS achieves computational speeds that are 6 $\times$ faster than those of default implementations.

The TAT-pulsar (Timing Analysis Toolkit for Pulsars) package is a specialized toolkit designed for handling the scientific intricacies of pulsar timing. It provides a suite of Python-based utilities and scripts that facilitate the analysis, processing, and visualization of pulsar data. By leveraging observational data from pulsars, along with the associated physical processes and statistical characteristics, TAT-pulsar integrates a series of useful tools and data analysis scripts specifically developed for both isolated pulsars and binary systems. This enables swift analysis and the detailed presentation of timing properties in the high-energy pulsar field. Developed and implemented completely independently from other pulsar timing software such as Stingray (ascl:1608.001) and PINT (ascl:1902.007), TAT-pulsar serves as a valuable cross-checking and supplementary tool for data analysis.

polarizationtools converts, analyzes, and simulates polarization data. The different python scripts (1) convert Stokes parameters into linear polarization parameters with proper treatment of the uncertainties and vice versa; (2) shift electric vector position angle (EVPA) data points in time series to account for the 180 degrees ambiguity; (3) identify rotations of the EVPA e.g. in blazar polarization monitoring data according to various rotation definitions; and (4) simulate polarization time series as a random walk in the Stokes Q-U plane.

ECLIPSR fully and automatically analyzes space based light curves to find eclipsing binaries and provide some first order measurements, such as the binary star period and eclipse depths. It provides a recipe to find individual eclipses using the time derivatives of the light curves, including eclipses in light curves of stars where the dominating variability is, for example, pulsations. Since the algorithm detects each eclipse individually, even light curves containing only one eclipse can (in principle) be successfully analyzed and classified. ECLIPSR can find eclipsing binaries among both pulsating and non-pulsating stars in a homogeneous and quick manner and process large amounts of light curves in reasonable amounts of time. The output includes, among other things, the individual eclipse markers, the period and time of first (primary) eclipse, and a score between 0 and 1 indicating the likelihood that the analyzed light curve is that of an eclipsing binary.

star_shadow automatically analyzes space based light curves of eclipsing binaries and provide a measurement of eccentricity, among other parameters. It measures the timings of eclipses using the time derivatives of the light curves, using a model of orbital harmonics obtained from an initial iterative prewhitening of sinusoids. Since the algorithm extracts the harmonics from the rest of the sinusoidal variability eclipse timings can be measured even in the presence of other (astrophysical) signals, thus determining the orbital eccentricity automatically from the light curve along with information about the other variability present in the light curve. The output includes, but is not limited to, a sinusoid plus linear model of the light curve, the orbital period, the eccentricity, argument of periastron, and inclination.

SkyLine generates mock line-intensity maps (both in 3D and 2D) in a lightcone from a halo catalog, accounting for the evolution of clustering and astrophysical properties, and observational effects such as spectral and angular resolutions, line-interlopers, and galactic foregrounds. Using a given astrophysical model for the luminosity of each line, the code paints the signal for each emitter and generates the map, adding coherently all contributions of interest. In addition, SkyLine can generate maps with the distribution of Luminous Red Galaxies and Emitting Line Galaxies.

2cosmos is a modification of Monte Python (ascl:1307.002) and allows the user to write likelihood modules that can request two independent instances of CLASS (ascl:1106.020) and separate dictionaries and structures for all cosmological and nuisance parameters. The intention is to be able to evaluate two independent cosmological calculations and their respective parameters within the same likelihood. This is useful for evaluating a likelihood using correlated datasets (e.g. mutually exclusive subsets of the same dataset for which one wants to take into account all correlations between the subsets).

cbeam is a Python/Julia package which models the propagation of guided light through slowly-varying few-mode waveguides using the coupled-mode theory (CMT). When compared with more general numerical methods for waveguide simulation, such as the finite-differences beam propagation method (FD-BPM), numerical implementations of the CMT can be much more computationally efficient. cbeam also provides a Pythonic class structure to define waveguides, with simple classes for directional couplers and photonic lanterns already provided. Finally, cbeam doubles as a finite-element eigenmode solver.

KCWIKit extends the official KCWI DRP (ascl:2301.019) with a variety of stacking tools and DRP improvements. The software offers masking and median filtering scripts to be used while running the KCWI DRP, and a step-by-step KCWI_DRP implementation for finer control over the reduction process. Once the DRP has finished, KCWIKit can be used to stack the output cubes via the Montage package. Various functions cross-correlate and mosaic the constituent cubes and the final stacked cubes are WCS corrected. Helper functions can then be used to deproject the stacked cube into lower-dimensional representations should the user desire.

Pynkowski computes Minkowski Functionals and other higher order statistics of input fields, as well as their expected values for different kinds of fields. This package supports Minkowski functionals, and maxima and minima distributions. Supported input formats include scalar HEALPix maps such as those used by healpy (ascl:2008.022) and polarization HEALPix maps in the SO(3) formalism. Pynkowski also supports various theoretical fields, including Gaussian (e.g., CMB Temperature or the initial density field), Chi squared (e.g., CMB polarization intensity), and spin 2 maps in the SO(3) formalism.

DistClassiPy uses different distance metrics to classify objects such as light curves. It provides state-of-the-art performance for time-domain astronomy, and offers lower computational requirements and improved interpretability over traditional methods such as Random Forests, making it suitable for large datasets. DistClassiPy allows fine-tuning based on scientific objectives by selecting appropriate distance metrics and features, which enhances its performance and improves classification interpretability.

kinematic_scaleheight uses MCMC methods to kinematically estimate the vertical distribution of clouds in the Galactic plane, including the least squares analysis of Crovisier (1978), an updated least squares analysis using a modern Galactic rotation model, and a Bayesian model sampled via MCMC as described in Wenger et al. (2024).

BTSbot automates real-time identification of bright extragalactic transients in Zwicky Transient Facility (ZTF) data. A multi-modal convolutional neural network, BTSbot provides a bright transient score to individual ZTF detections using their image data and 25 extracted features. The package eliminates the need for daily visual inspection of new transients by automatically identifying and requesting spectroscopic follow-up observations of new bright transient candidates. BTSbot recovers all bright transients in our test split and performs on par with human experts in terms of identification speed (on average, ∼1 hour quicker than scanners).

Poke (pronounced /poʊˈkeɪ/ or po-kay) uses commercial ray tracing APIs and open-source physical optics engines to simultaneously model scalar wavefront error, diffraction, and polarization to bridge the gap between ray trace models and diffraction models. It operates by storing ray data from a commercial ray tracing engine into a Python object, from which physical optics calculations can be made. Poke provides two propagation physics modules, Gaussian Beamlet Decomposition and Polarization Ray Tracing, that add to the utility of existing scalar diffraction models. Gaussian Beamlet Decomposition is a ray-based approach to diffraction modeling that integrates physical optics models with ray trace models to directly capture the influence of ray aberrations in diffraction simulations. Polarization Ray Tracing is a ray-based method of vector field propagation that can diagnose the polarization aberrations in optical systems.

fkpt computes the 1-loop redshift space power spectrum for tracers using perturbation theory for LCDM and Modified Gravity theories using "fk"-Kernels. Though implemented for the Hu-Sawicky f(R) modified gravity model, it is straightforward to use it for other models.

The MINDS hybrid pipeline is based on the JWST pipeline and routines from the VIP package (ascl:1603.003) for the reduction of JWST MIRI-MRS data. The pipeline compensates for some of the known weaknesses of the official JWST pipeline to improve the quality of spectrum extracted from MIRI-MRS data. This is done by leveraging the capabilities of VIP, another large data reduction package used in the field of high-contrast imaging.

The front end of the pipeline is a highly automated Jupyter notebook. Parameters are typically set in one cell at the beginning of the notebook, and the rest of the notebook can be run without any further modification. The Jupyter notebook format provides flexibility, enhanced visibility of intermediate and final results, more straightforward troubleshooting, and the possibility to easily incorporate additional codes by the user to further analyze or exploit their results.

s4cmb (Systematics For Cosmic Microwave Background) studies the impact of instrumental systematic effects on measurements of CMB experiments based on bolometric detector technology. s4cmb provides a unified framework to simulate raw data streams in the time domain (TODs) acquired by CMB experiments scanning the sky, and to inject in these realistic instrumental systematics effect.

pycorr wraps two-point counter engines such as Corrfunc (ascl:1703.003) to estimate the correlation function. It supports theta (angular), s, s-mu, rp-pi binning schemes, analytical two-point counts with periodic boundary conditions, and inverse bitwise weights (in any integer format) and (angular) upweighting. It also provides MPI parallelization and jackknife estimate of the correlation function covariance matrix.

FitCov estimates the covariance of two-point correlation functions in a way that requires fewer mocks than the standard mock-based covariance. Rather than using an analytically fixed correction to some terms that enter the jackknife covariance matrix, the code fits the correction to a mock-based covariance obtained from a small number of mocks. The fitted jackknife covariance remains unbiased, an improvement over other methods, performs well both in terms of precision (unbiased constraints) and accuracy (similar uncertainties), and requires significant less computational power. In addition, FitCov can be easily implemented on top of the standard jackknife covariance computation.

LtU-ILI (Learning the Universe Implicit Likelihood Inference) performs machine learning parameter inference. Given labeled training data or a stochastic simulator, the LtU-ILI piepline automatically trains state-of-the-art neural networks to learn the data-parameter relationship and produces robust, well-calibrated posterior inference. The package comes with a wide range of customizable complexity, including posterior-, likelihood-, and ratio-estimation methods for ILI, including sequential learning analogs, and various neural density estimators, including mixture density networks, conditional normalizing flows, and ResNet-like ratio classifiers. It offers fully-customizable, exotic embedding networks, including CNNs and Graph Neural Networks, and a unified interface for multiple ILI backends such as sbi, pydelfi, and lampe. LtU-ILI also handles multiple marginal and multivariate posterior coverage metrics, and offers Jupyter and command-line interfaces and a parallelizable configuration framework for efficient hyperparameter tuning and production runs.

Pylians3 (Python3 libraries for the analysis of numerical simulations) provides a Python 3 version of Pylians (ascl:1811.008), which analyzes numerical simulations (both N-body and hydrodynamic); parts of the codebase are also written in cython and C. It computes density fields, power spectra, bispectra, and correlation functions, identifies voids, and populates halos with galaxies using an HOD. Pylians3 also applies HI+H2 corrections to the output of hydrodynamic simulations, make 21cm maps, computes DLAs column density distribution functions, and can plot density fields and make movies.

URecon reconstructs the initial conditions of N-body simulations from late time (e.g., z=0) density fields. This simple UNET architecture is implemented in TensorFlow and requires Pylians3 (ascl:2403.012) for measuring power spectrum of density fields. The package includes weights trained on Quijote fiducial cosmology simulations.

OneLoopBispectrum computes the one-loop bispectrum of galaxies in redshift space. It computes and simplifies the bispectrum kernels using Mathematica; this is cosmology-independent. The code also computes the full and flattened bispectrum templates, given the pre-computed integration kernels. OneLoopBispectrum uses Mathematica to read in and combine the bispectrum templates, and Python to interpolate and extract the one-loop bispectrum.

CLASS-PT modifies the CLASS (ascl:1106.020) code to compute the non-linear power spectra of dark matter and biased tracers in one-loop cosmological perturbation theory, for both Gaussian and non-Gaussian initial conditions. CLASS-PT can be interfaced with the MCMC sampler MontePython (ascl:1805.027) using the (new and improved) custom-built likelihoods found here.

The DensityFieldTools toolset manipulates density fields and measures power spectra and bispectra using a very simple interface. After loading a density field, it computes the power spectrum and the bispectrum for a desired binning. The bispectrum estimator also automatically computes the power spectrum for the chosen binning, to facilitate, for example, shot-noise subtraction. DensityFieldTools also provides a quick way to measure (cross-)power spectra directly from density fields.

Obsplanning is a suite of tools to help plan astronomical observations from ground-based observatories, for traditional single-site as well as multi-station (VLBI) observing. Conveniently determine observability of objects in the sky from your observatory, and produce plots to help you prepare for your observations over the course of a session. Celestial source coordinates (including solar system objects) can be queried or created, and transformed. Calibrator or reference sources can be selected by proximity, and slew order can be optimized to save valuable telescope time. Plots and visualizations can be easily made to chart source elevation and transits, source proximity to the Sun and Moon, concurrent 'up time' of sources at multiple sites (for VLBI or tandem observations), 'dark time' at a telescope site for a given year, finder plots made from real images (with options to query online databases), and more.

This package provides some utilities for binning, normalizing colors, wrangling tick marks, and more, in symmetric logarithm space. That is, for numbers spanning positive and negative values, working in log scale with a transition through zero, down to some threshold. This can be quite useful for representing data that span many scales like standard log-space, but that include values of zero (that might be returned by physical measurement) or even negative values (for example offsets from some reference, or things like temperatures). This package provides convenient functions for creating 1D and 2D histograms and symmetric log bins, generating logspace-like arrays through zero and managing matplotlib major and minor ticks in symlog space, as well as bringing symmetric log scaling functionality to plotly.

PIPE (PSF Imagette Photometric Extraction) extracts PSF (point-spread function) photometry from data acquired by the space telescope CHEOPS (CHaracterisation of ExOPlanetS). Advantages of PSF photometry over standard aperture photometry include reduced sensitivity to contaminants such as background stars, cosmic ray hits, and hot/bad pixels. For CHEOPS, an additional advantage is that photometry can be extracted from an imagette, a small window around the target that is downlinked at a shorter cadence than the larger-sized subarray used for aperture photometry. These advantages make PIPE particularly well suited for targets brighter or fainter than the nominal G = 7-11 mag range of CHEOPS, i.e. where short-cadence imagettes are available (bright end) or when contamination becomes a problem (faint end). Within the nominal range, PIPE usually offers no advantage over the standard aperture photometry.