Search Results for "kd tree"

libadb is an astronomical catalog management library desgined to provide fast access to astronomical object data.

Implementation of Point Symmetry-based Automatic Clustering Approach Using Differential Evolution Using bug fixed KD tree nearest neighbor search from https://github.com/CristianDallos/kmeansclustering. (It is also modified to search for multiple nearest points instead only for one.) Used those academic works for algorithm implementation: http://cs.cug.edu.cn/teacherweb/gwy/Publication/ISICA-09.pdf http://www.isical.ac.in/~sanghami/ieeetkde_cameraready.pdf http://www.softcomputing.net/smca-paper1.pdf 1. ...

Extreem Reader for Cosmological simulation data P-Gadet2 and ART codes.It conserves memory by using memory mapped files. Contains some basic analysis tools such a KD-Tree, Minimum Spanning Tree, and rendering module using PMViewer algorithm.

Spatial is a generic header-only C++ library providing multi-dimensional in-memory containers, iterators and functionals. Spatial is a general purpose library with good overall performance thanks to its self-balancing k-d tree containers. Spatial makes it easy to search for nearest neighbors, orthogonal ranges, or perform ordering of objects in any number of dimensions. Spatial's design borrows heavily from the STL, and strives to follow it's elegant interface; therefore Spatial's containers, iterators and functionals integrate well with the STL. Additionally, Spatial's self-balancing kd-tree based containers, capable of object insertion and removal in fractional amortized time, prevent worst-case computational time on most operations. ...

Fast Matrix for Java (fm4j) is a general-purpose matrix utility library for computing with dense matrices. fm4j encapsulated different underlying implementations and select the optimal one in run-time depending on the size of the input matrix. Moreover, fm4j employs Java (Tm) Concurrency to take advantage of the computation power of multi-cor processors.

This is a simple ray tracing algorithm implementation. The same algorithm is implemented on CPU and GPU with CUDA, and the performance of the GPU based ray tracing is much better than the naive CPU one. cuda and directx sdk are required.

We have implemented a raytracer in both CTM and DirectX (through brook). On modern architectures the raytracer gets between 10 and 20 million rays per second on 300,000 polygon scenes.... and even more on quake levels and other scenes.

a C++ template container implementation of k-dimensional space sorting based on the kd-tree data structure. THIS PROJECT HAS MOVED TO ALIOTH.DEBIAN.ORG