[NeurIPS'21] Shape As Points: A Differentiable Poisson Solver

A flexible framework for solving PDEs with modern spectral methods.

High performance computational platform in Python for the spectral Galerkin method

Implementation of Directional Graph Networks in PyTorch and DGL

Comparison of various numerical methods for computational fluid dynamics

Systems Neuroscience Computing in Python: user-friendly analysis of large-scale electrophysiology data

GARNET: Reduced-Rank Topology Learning for Robust and Scalable Graph Neural Networks

[WWW 2023] "Addressing Heterophily in Graph Anomaly Detection: A Perspective of Graph Spectrum" by Yuan Gao, Xiang Wang, Xiangnan He, Zhenguang Liu, Huamin Feng, Yongdong Zhang

Pseudospectral Kolmogorov Flow Solver

ManifoldEM Python suite

PySpectral is a Python package for solving the partial differential equation (PDE) of Burgers' equation in its deterministic and stochastic version.

Runge-Kutta adaptive-step solvers for nonlinear PDEs. Solvers include both exponential time differencing and integrating factor methods.

Fast and flexible two- and three-point correlation analysis for time series using spectral methods.

📈 Common API (C++ and Python) for Fast Fourier Transform HPC libraries (publish-only mirror)

Physically-informed model discovery of systems with nonlinear, rational terms using the SINDy-PI method. Contains functionality for spectral filtering/differentiation.

Working with numerical grids made easy.

Radio-wave transfer matrix model for glacier ice

Software implementation for paper Heuristic Analysis of Manifolds from Simulated Cryo-EM Ensemble Data

Under the supervision of Prof. Matthew Browning, I was tasked to study the creation of a spectral solver for Boussinesq approximated fluid dynamics and then, using Dedalus, to study the effects of rotation and stratification on turbulent convection.