Flax provides a flexible end-to-end user experience for JAX users; its NNX is a simplified API that creates, inspects, debugs, and analyzes neural networks in JAX. It has first class support for Python reference semantics, enabling users to express their models using regular Python objects. Flax NNX is an evolution of the previous Flax Linen API.

picasso makes predictions for the thermodynamic properties of the gas in massive dark matter halos from gravity-only cosmological simulations. It combines an analytical model of gas properties as a function of gravitational potential with a neural network predicting the parameters of said model. Written in Python, it combines an implementation of the gas model based on JAX (ascl:2111.002) and Flax (ascl:2504.026), and models that have been pre-trained to reproduce gas properties from hydrodynamic simulations.

jaxoplanet is a functional-programming-forward implementation of many features from the exoplanet and starry packages built on top of JAX (ascl:2111.002). It includes fast and robust implementations of many exoplanet-specific operations, including solving Kepler’s equation, and computing limb-darkened light curves. Built on top of JAX, jaxoplanet has first-class support for hardware acceleration using GPUs and TPUs, and integrates seamlessly with modeling tools such as NumPyro and Flax (ascl:2504.026).

The GNU Radio toolkit provides signal processing blocks to implement software radios. A software radio performs signal processing in software instead of using dedicated integrated circuits in hardware. The benefit is that since software can be easily replaced in the radio system, the same hardware can be used to create many kinds of radios for many different communications standards. GNU Radio can be used with readily-available low-cost external RF hardware to create software-defined radios and to simulate wireless communications.

The highly optimized Kotekan framework processes streaming data. Written in a combination of C/C++, it is primarily designed for use on radio telescopes and was originally developed for the CHIME project. It is similar to radio projects such as GNUradio (ascl:2504.029) or Bifrost (ascl:1711.021), though has a greater focus on efficiency and throughput. Kotekan is conceptually straightforward: data is carried through the system in a series of ring buffer objects, which are connected by processing blocks which manipulate the data, and optional metadata structures can be passed alongside the streaming data.

TempoNest performs a Bayesian analysis of pulsar timing data, which allows for the robust determination of the non-linear pulsar timing solution simultaneously with a range of additional stochastic parameters. This includes both red spin noise and dispersion measure variations using either power law descriptions of the noise, or through a model-independent method that parameterizes the power at individual frequencies in the signal. It uses the Bayesian inference tool MultiNest (ascl:1109.006) to explore the joint parameter space, while using Tempo2 (ascl:1210.015) as a means of evaluating the timing model. TempoNest allows for the analysis of additional stochastic signals beyond the white noise described by the TOA error bars that may be present in the data.

DMCalc estimates the Dispersion Measure (DM) of wide-band pulsar data in psrfits format. It uses PSRCHIVE (ascl:1105.014) tools to get ToAs and then uses TEMPO2 (ascl:1210.015) for DM fitting. A median absolute deviation (MAD) based ToA rejection algorithm is implemented in the code to remove large outlier ToAs using Huber Regression. Although the code has been used for analyzing uGMRT wide-band data, DMCalc can in principle be used for any pulsar dataset.

Vela.jl performs Bayesian pulsar timing and noise analysis. It supports narrowband and wideband TOAs along with most commonly used pulsar timing models. The code provides an independent, efficient, and parallelized implementation of the full nonlinear pulsar timing and noise model and includes a Python binding (pyvela). One-time operations such as data file input, clock corrections, and solar system ephemeris computations are performed by pyvela with the help of the PINT (ascl:1902.007) pulsar timing package.

JOFILUREN analyzes and de-noises scientific data and is useful for studying and reducing the physical effects of particle noise in particle-mesh computer simulations. It uses wavelets, which can efficiently remove noise from cosmological, galaxy and plasma N-body simulations. Written in Fortran, the code is portable and can be included in grid-based N-body codes. JOFILUREN can also be applied for removing noise from standard data, such as signals and images.

SHELLFISH (SHELL Finding In Spheroidal Halos) finds the splashback shells of individual halos within cosmological simulations. It uses a command line toolchain to produce human-readable catalogs. It requires a configuration file that describes the layout of the particle snapshots and halo catalog and which halos to measure the splashback shell for; once that is provided, Shellfish takes care of the rest. It supports numerous particle catalog types, including gotetra, Gadget-2, and Bolshoi, all text column-based halo catalogs, and consistent-trees merger trees.

show_cube displays the results of reducing, aligning and combining near-infrared integral field spectroscopy with the Gemini Observatory NIFS (Near-infrared Integral Field Spectrometer) instrument. Image slices are extracted from the raw data frames to make the input datacube. The code site also provides a tarfile containing all the raw NIFS FITS-format files for the observations of high-redshift radio galaxies 3C230, 3C294, and 4C+41.17, the last of which are reported, together with line-strengths using the MAPPINGS III (ascl:1306.008) shock models.

astromorph performs an automatic classification of astronomical objects based on their morphology using machine learning in a self-supervised manner. Written in Python, the pipeline is an implementation for astronomical images in FITS-format files of the Boot-strap Your Own Latents (BYOL; Grill et al. 2020) method, which does not require labelling of the training data.

We introduce COBRA (Cosmology with Optimally factorized Bases for Rapid Approximation), a novel framework for rapid computation of large-scale structure observables. COBRA separates scale dependence from cosmological parameters in the linear matter power spectrum while also minimising the number of necessary basis terms, thus enabling direct and efficient computation of derived and nonlinear observables. Moreover, the dependence on cosmological parameters is efficiently approximated using radial basis function interpolation. We apply our framework to decompose the linear matter power spectrum in the standard LCDM scenario, as well as by adding curvature, dynamical dark energy and massive neutrinos, covering all redshifts relevant for Stage IV surveys. With only a dozen basis terms, COBRA reproduces exact Boltzmann solver calculations to 0.1% precision, which improves further to 0.02% in the pure LCDM scenario. Using our decomposition, we recast the one-loop redshift space galaxy power spectrum in a separable minimal-basis form, enabling $\sim 4000$ model evaluations per second at 0.02% precision on a single thread. This constitutes a considerable improvement over previously existing methods (e.g., FFTLog) opening a new window for efficient computations of higher loop and higher order correlators involving multiple powers of the linear matter power spectra. The resulting factorisation can also be utilised in clustering, weak lensing and CMB analyses. Our implementation is publicly available at https://github.com/ThomasBakx/cobra.

arctic_weather reports analysis of meteorological data recorded from High Arctic weatherstations (called Inuksuit) deployed on coastal mountains north of 80 degrees on Ellesmere Island Canada from 2006 through 2009, along with clear-sky fractions from horizon-viewing sky-monitoring cameras. The code calculates solar and lunar elevations, and so allows correlation of polar nighttime to the development of prevailing thermal inversion conditions in winter, and statistical comparison to other optical/infrared observatory sites.

allsky performs photometry of Polaris with the Polar Environment Atmospheric Research Laboratory (PEARL) All-Sky Camera (PASI) to report transparency measurements, with comparison to conditions at other observatories worldwide. The code site provides a tarfile of PASI data obtained near Eureka, on Ellesmere Island Canada in darktime of 2008/09 and 2009/10 along with associated meteorological data. The code employs a simple atmospheric thermal inversion model, with a power-law fit to ice-crystal attenuation, allowing direct comparison of PEARL dark-time photometric-sky statistics to mid-latitude sites such as Maunakea.

arctic_mass_dimm reduces data from the Multi-Aperture Seeing Sensor (MASS) and Differential Image Motion Monitor (MASS) obtained from the Polar Environment Atmospheric Research Laboratory (PEARL), reporting seeing conditions, and comparing to other observatories. The code site provides a tarfile of all MASS and DIMM data obtained near Eureka, on Ellesmere Island Canada in 2011/12 along with associated meteorological data. The code employs a simple two-component atmospheric model to allow comparison of PEARL to mid-latitude sites such as Maunakea.

infrared_comparison compares the downwelling infrared radiation, or sky spectral brightness, of arctic/antarctic astronomical observing sites with the best mid-latitude mountain sites. The code site provides a tarfile of Fourier-transform spectra from 3.3 microns 20 microns, obtained near Eureka, on Ellesmere Island Canada, along with meteorological data. The code can compare these via an atmospheric thermal-inversion model to reported values for South Pole and other mid-latitude sites, such as Maunakea.

Sapphire++ is an open-source code designed to numerically solve the Vlasov–Fokker–Planck equation for astrophysical applications. Sapphire++ employs a numerical algorithm based on a spherical harmonic expansion of the distribution function, expressing the Vlasov–Fokker–Planck equation as a system of partial differential equations governing the evolution of the expansion coefficients. The code utilises the discontinuous Galerkin method in conjunction with implicit and explicit time stepping methods to compute these coefficients, providing significant flexibility in its choice of spatial and temporal accuracy.

This modular Python-based pipeline provides tools for computing background cosmological quantities and Fourier-space power spectra for multiple tracers of large-scale structure, such as galaxies and 21cm intensity maps. It is designed for multitracer Fisher forecasting in both the linear regime and nonlinear scales using HALOFIT. The pipeline enables forecasts of cosmological parameters such as f_NL, fσ8, and tracer bias parameters. Its flexible architecture includes independently callable modules for the Hubble parameter, comoving distance, growth functions, matter power spectrum, transfer functions, and cross-power spectrum combinations. The code supports both theoretical survey design and nonlinear parameter estimation, making it suitable for a wide range of cosmological analyses.

This modular Python-based pipeline provides tools for computing background cosmological quantities and Fourier-space power spectra for multiple tracers of large-scale structure, such as galaxies and 21cm intensity maps. It is designed for multitracer Fisher forecasting in both the linear regime and nonlinear scales using HALOFIT. This version reflects a cleaned folder structure, updated filenames, and improved modularity. It supports forecasting of parameters such as f_NL, fσ8, and bias terms.

This modular Python-based pipeline provides tools for computing background cosmological quantities and Fourier-space power spectra for multiple tracers of large-scale structure, such as galaxies and 21cm intensity maps. It is designed for multitracer Fisher forecasting in both the linear regime and nonlinear scales using HALOFIT. The pipeline enables forecasts of cosmological parameters such as f_NL, fσ8, and tracer bias parameters. Its flexible architecture includes independently callable modules for the Hubble parameter, comoving distance, growth functions, matter power spectrum, transfer functions, and cross-power spectrum combinations. The code supports both theoretical survey design and nonlinear parameter estimation, making it suitable for a wide range of cosmological analyses.

This modular Python-based pipeline provides tools for computing background cosmological quantities and Fourier-space power spectra for multiple tracers of large-scale structure, such as galaxies and 21cm intensity maps. It is designed for multitracer Fisher forecasting in both the linear regime and nonlinear scales using HALOFIT. The pipeline enables forecasts of cosmological parameters such as f_NL, fσ8, and tracer bias parameters. Its flexible architecture includes independently callable modules for the Hubble parameter, comoving distance, growth functions, matter power spectrum, transfer functions, and cross-power spectrum combinations. The code supports both theoretical survey design and nonlinear parameter estimation, making it suitable for a wide range of cosmological analyses.