2026
27 projects
Numba Optimized Implementations for scikit bio Performance Critical Functions
scikit-bio relies on Cython extensions for performance-critical distance and ordination operations, but these cannot accept GPU arrays. This project...
Lumen + Xarray Integration
Lumen handles tabular data through DuckDB but cannot work with N-dimensional scientific formats like NetCDF, Zarr, and HDF5. This project brings...
Serialization Of Map and Model Classes into ASDF format
Gammapy currently serializes its data products across multiple data formats, which creates a fragmented workflow with poor metadata handling. This...
pvlib: Standardizing Diffuse Irradiance Components and Extending ModelChain's Optical Loss Modeling
This project aims to improve the handling of transposed diffuse irradiance and optical loss modeling in the pvlib-python package. Currently, pvlib’s...
Large Scale HEOM Solver Using PETSc
This project aims to scale QuTiP's HEOM solver beyond single-node limitations by integrating PETSc to distribute ODE integration across multiple MPI...
Predictively Oriented Posteriors in PyMC
Standard Bayesian posteriors degrade under model misspecification by collapsing to point estimates. This project implements Predictively Oriented...
Native Julia ODE, SDE, DAE, DDE, and (S)PDE Solvers
OrdinaryDiffEq.jl contains over 18 IMEX Runge–Kutta implementations that share identical stepping logic but have bespoke per-method perform_step!...
Reduced Order Modelling with Neural Operators in Gridap.jl
THE PROBLEM: Solving partial differential equations (PDEs) for parametrical studies is a fundamental part of scientific research, but traditional...
Distributed Linear Solvers in LinearSolve.jl
This project will add distributed linear solver support to LinearSolve.jl by expanding and hardening PETSc.jl integration. The goal is to let SciML...
AiiDA Copilot: AiiDA Natural Language Interface Using Multi-Agent AI
AiiDA workflows currently require manual Python scripting. This project builds a multi-agent AI system that lets scientists interact with AiiDA...
Stateless API for Scalable and Composable Inference in sbi
This project redesigns sbi's potential function API to eliminate stateful behavior, inconsistent IID handling, and device management bugs. The new...
Parallelizing r.proj and Raster Processing Modules in GRASS
r.proj and r.param.scale are widely used GRASS raster modules that run entirely on a single core, leaving modern multi-core hardware idle. This...
Supporting Input Convex Neural Networks in MathOptAI.jl
MathOptAI.jl is a Julia package for embedding machine learning models into optimization problems built using JuMP. These embeddings work as...
QuTiP: GPU-enabled circuit simulation in qutip-qip
The current gate-level circuit simulator in qutip-qip utilizes a full sparse Qobj representation of each gate, expanding local gates to the Hilbert...
DevOps improvement and standardization
This project focuses on improving the development and testing infrastructure across the ArviZ ecosystem (arviz-base, arviz-stats, and arviz-plots)....
Add Spatio-Temporal Datasets Support to GRASS Data Catalog GUI
Currently, the GRASS Data Catalog only queries standard spatial maps and bypasses the temporal database and STDS's entirely. As a result, Space Time...
DeepForest P_2: Recovering Historical Image Data Using Automated OrthoRegistration and ImageMatching
The U.S. Forest Service's Aerial Detection Survey (ADS) program has produced tens of thousands of hand-drawn polygons marking forest health damage...
Neural Network Builder API Refactor
The sbi (Simulation-Based Inference) library currently constructs neural density estimators via opaque factory functions that accept loose keyword...
PySal: Implementation of Geographically Weighted Matrix Decomposition Algorithms in gwlearn
Standard PCA assumes a globally stationary covariance structure, missing spatial heterogeneity in geographic data. This project implements...
Panel / HoloViews MCP Integration
Panel and HoloViews are powerful Python visualization libraries but they have no MCP integration today. AI agents can write visualization code but...
Trainite: A Reusable Toolbox for Training Language Models
Every time a researcher wants to train a language model, they end up rewriting the same infrastructure: data loading, optimizer setup, training...
Succinct Data Structure for Efficient Operations on Trees in Scikit-bio
Scikit-bio offers a great toolkit for working with biological data that has a tree-like structure. Although the current implementation, TreeNode, can...
SciML Comprehensive Non Uniform Grid Support for MethodOfLines
This project aims to deliver comprehensive, end-to-end non-uniform grid support for MethodOfLines.jl. Currently, basic advection schemes lack...
DeepForest: Recovering computer vision annotations from historical airborne imagery monitoring
After the 2010 Deepwater Horizon oil spill, airborne surveys produced 18,304 images with 2.6 million birds individually marked using a point-counting...
Matplotlib Overlay Layer API for Smooth Interactive Tools
I propose to implement the Overlay Layer API for Matplotlib’s interactive backends. Currently, tools like the Cursor widget force a full figure...
Parallelizing loo
The `loo` package is a widely used R package in the Stan ecosystem which exposes a clean API to cross-validate (CV) Bayesian models. Currently, some...
PyMC Streaming Variational Inference for Large Datasets
PyMC's variational inference stack — including ADVI and Pathfinder — assumes all data fits in memory. Real-world applications such as financial tick...