CORTEXA
← Browse
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-24Cited by 0

Isurwars/Correlation: Correlation 3.7.0

Isaías Rodríguez Aguirre, Mineralwater Xu

[3.7.0] - 2026-07-24 Added Multi-Vendor GPU Acceleration Framework: Implemented SYCL-based GPU acceleration for distance, $S(Q)$, and Steinhardt parameter calculators (GPUDistanceCalculator, GPUSQCalculator, GPUSteinhardtCalculator) with runtime fallback and device detection. GPU Portability Abstraction Layer: Added GPUPortability layer featuring unified HIP and CUDA compatibility wrappers for high-throughput pairwise distance calculations. Machine Learning Interatomic Potentials (MLIP): Integrated MLIPCalculator and MLIPInterface for machine learning interatomic potential calculations. Modular SIMD Acceleration Framework: Introduced modular SIMD acceleration with AVX2, AVX512, and scalar fallbacks, including vectorized dot_block kernels and templatized SIMD data structures (Vector3SIMD). Configurable Floating-Point Precision: Replaced hardcoded double precision with real_t type aliases across all core calculators, readers, plotters, and test suites, enabling runtime/compile-time single or double precision builds via USE_SINGLE_PRECISION. Cross-Platform PyPI Wheel Automation: Added repair_wheel.py helper script to automate oneTBB dependency vendoring and rpath patching for Python wheel builds. Changed & Refactored Parameter Struct Standardization: Refactored DADCalculator, VDOSCalculator, ScaleBins, and RDFCalculator call interfaces to use dedicated parameter structures (ScaleBinsParams, VDOSParams, RDFNormalizationParams) for type-safe dispatch. Graph Edge Representation: Introduced unified Edge structure and overloaded addDirectedEdge in DistanceCalculator for cleaner AST and graph edge insertion. Codebase Modernization & Linter Compliance: Replaced reinterpret_cast with C++20 std::bit_cast, standardized PRNG seeding in tests, and eliminated all NOLINT/NOLINTNEXTLINE inline suppressions across source files and documentation. Project-Wide Formatting: Applied uniform clang-format styling across the entire project codebase. Packaging & Build System: Enforced static builds for voro++, standardized installation prefixes, and updated .desktop entry paths. Performance & Optimization Voronoi Topology Optimization: Accelerated VoronoiCalculator by caching thread-local buffers, filtering parasitic edges in Voronoi signatures, and optimizing particle ordering in periodic boundary containers. Bounds-Checked Operator Efficiency: Replaced unchecked raw array indexing with bounds-safe .at() calls in mathematical structures, while keeping direct array indexing in performance-critical Matrix3 operator kernels. Bit-Canonical Sampling: Implemented portable bit-canonical random sampling and modularized linear slope fitting in analysis modules. Fixed Voronoi Coordinate Normalization: Corrected floating-point normalization and hyperuniformity fitting bounds for Voronoi coordinate calculations. Cross-Platform Stat Header: Corrected header paths for filesystem stat structure to ensure cross-platform header compatibility. Precision Test Assertions: Migrated floating-point test assertions to custom IsRealEq matchers for reliable multi-precision test validation. Full Changelog: https://github.com/Isurwars/Correlation/compare/v1.0.0...v3.7.0

View free PDFSource page

Related papers

openalexZenodo (CERN European Organization for Nuclear Research)2026-07-24

Isurwars/Correlation: Correlation 3.7

Isaías Rodríguez Aguirre, Mineralwater Xu

[3.7.0] - 2026-07-24 Added Multi-Vendor GPU Acceleration Framework: Implemented SYCL-based GPU acceleration for distance, $S(Q)$, and Steinhardt parameter calculators (GPUDistanceCalculator, GPUSQCalculator, GPUSteinhardtCalculator) with runtime fallback and device detection. GPU…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-26

Isurwars/Correlation: Correlation 3.7.2

Isaías Rodríguez Aguirre, Mineralwater Xu

[3.7.2] - 2026-07-26 Fixed Native File Dialog Linkage: Enforced static library compilation (BUILD_SHARED_LIBS=OFF) for nativefiledialog-extended (nfd) on Linux to resolve runtime libnfd.so.1 shared library loading errors. [3.7.0] - 2026-07-24 Added Multi-Vendor GPU Acceleration F…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-23

SPADE OC2 Mediterranean Trefoil crop dataset (T1) - VITAFields – Visual Intelligence for Targeted Agriculture #4

George Kougianos, Athanasios Kerasiotis, Maria Petsa, Emmanouil Taoulai

This dataset release represents Part 4 of the comprehensive young trefoil crop agricultural analysis project. While Part 1, Part 2 and Part 3 provided the raw image captures and camera parameters and machine-learning-ready image tiles. Part 4 delivers fully processed, georeferenc…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-25

QSM-CI method: INR-QSM (v1)

M. Zhang, Feng R., Li Z., Feng J., Wu Q, Zhang Z., et al.

INR-QSM — a subject-specific UNSUPERVISED deep-learning dipole inversion using an implicit neural representation. No pretrained weights: a sine-activated coordinate MLP (SIREN) is OPTIMIZED per-subject so that the susceptibility it represents, pushed through the QSM dipole forwar…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-23

Data from: Automated Ethogram Elaboration: A Cross-species Deep-learning Model Deployed On-board Enables Acceleration-triggered Capture of Diverse Behavioural Video

Tsuneari Kuroiwa

This record contains field-deployment datasets for two seabird species, streaked shearwaters and black-tailed gulls, used in the paper “Automated Ethogram Elaboration: A Cross-species Deep-learning Model Deployed On-board Enables Acceleration-triggered Capture of Diverse Behaviou…

View free PDFSource page
openalexZenodo (CERN European Organization for Nuclear Research)2026-07-25

QSM-CI method: BFRnet (v1)

Xuanyu Zhu, Yang Gao, Feng Liu, Stuart Crozier, Hongfu Sun

Deep-learning background field removal (BFRnet): a 3D dual-frequency octave-convolution U-net trained to predict the background field of the brain — including brains with significant pathological susceptibility sources (haemorrhage, calcification). Consumes the total field (ppm)…

View free PDFSource page