highrisk的博客

Arbitrary Length FFT ------------------------------------------------------------------------ NOTE : This is copyrighted material, NOT public domain. See below. ------------------------------------------------------------------------ Contents: This packet contains the C source for a mixed-radix FFT routine. It performs a fast discrete Fourier transform (FFT) of a complex sequence, x, of an arbitrary length, n. The output, y, is also a complex sequence of length n. y[k] = sum(x[m]*exp(-i*2*pi*k*m/n), m=0..(n-1)), k=0,...,(n-1) The largest prime factor of n must be less than or equal to the constant, maxPrimeFactor defined in mixfft.c. The input/output variables are each stored in two arrays comprising the real part and the imaginary part respectively. The routine is accompanied by a demo program, fftbench.c, that demonstrates the numerical capabilities. It measures the execution time as well. If you wish to test the capabilities of another FFT routine, the fftbench.c is easy to modify. ------------------------------------------------------------------------ Filelist: mixfft.c 21.120 001030 The C source for the mixed-radix FFT. fftbench.c 8.724 001030 The C source for FFT benchmark program. fftbench.exe 62.976 001030 The PC executable. readme.txt This file. ------------------------------------------------------------------------ Platforms: PC: The included fftbench.exe was generated using the Microsoft Visual C++ compiler with the following compile options: /nologo /Gs /G2 /W4 /AH /Ox /D "NDEBUG" /D "_DOS" /FR It was tested on a 50MHz 486DX. Please refer to fftbench.c for some benchmarks. (jjn, March 1996) Other: Your experience with other platforms are warmly welcomed. Please e-mail me. ------------------------------------------------------------------------ Conditions: The source code in this packet is copyrighted material. Non-commercial use of the source code is free. A $100 fee must be paid if used commercially. Please contact me at jjn@get2net.dk and register your copy. The $100 fee includes up to one hour of assistance related to your use of the code. A trial period of 14 days is allowed. If the code is used for professional (paid) research and development for non-profit organisations like universities a reduced fee of $20 must be paid. The commercial license allows you to include the compiled code in a product or to use the code on a regular basis. You are however NOT allowed to sell the source code. Distribution of the complete unaltered package, including this file, is free. This includes commercial CD's. ------------------------------------------------------------------------ Author: Jens Joergen Nielsen For non-commercial use only. Bakkehusene 54 A $100 fee must be paid if used 2970 Hoersholm commercially. Please contact. DENMARK mailto:jjn@get2net.dk All rights reserved. October 2000. Check my homepage for updates and FAQ's http://home.get2net.dk/jjn ------------------------------------------------------------------------ Release notes: V.0.1 950301 Initial MIXFFT release. V.0.3 960317 Input/output variables have separate arrays for real and imag part. The speed of prime factors larger than 7 is more than doubled. V.0.4 980103 e-mail address changed. V.0.5 001030 Memory allocation in FFTBench changed, now calloc is used. Error reporting when primefactor is larger than maxPrimeFactor has been corrected. ------------------------------------------------------------------------

Windows系统高精度计时器 原理：联合使用QueryPerformanceFrequency和QueryPerformanceCounter获取高精度时间； 平台：vc++6/vc++2008

一个非常高效基于Kd-tree数据结构的2D、3D的近邻查询算法，原作者John Tsiombikas，使用C++封装并给了测试测序。

FasterCSV is intended as a replacement to Ruby's standard CSV library

boost::lexical_cast用法示例，包含数值转字串，字串转数值以及相应的异常处理代码

含GL.H、GLAUX.H、GLEXT.H、GLU.H、GLUT.H和WGLEXT.H，opengl/opengl32.lib、GLAUX.LIB、glu/glu32.lib、glut/glut32.lib以及相应的DLL文件，附VC环境设置说明文档。

VC++数据结构经典算法源码 二分查找1.c 二分查找2.c 二叉树.c [其它] [单元加] 单循环链表.c 单链表.c 图.c [字符] 定长串.c [小写数字转为大写数字] 带头结点双链循环线性表.c [底层编程] [效验算法] [数学问题] [数据结构] [数组] [文件程序] [求进制] [汉诺塔] [硬币情况] [逆阵] 链串.c 链栈.c 链队列.c [问题算法] 顺序栈.c 顺序表.c 顺序队列.c

A high-performance multi-dimensional C++ array class

GPU acceleration for STL algorithms Thrust提供了对C++ STL（主要是vector及相应算法、函数等）的GPU封装，而libcxxgpu基于Thrust的STL模板库的封装（如list、set等）。最新的版本提供对 gcc v4.5.1和v4.6.1的支持。

知名的WAP浏览器，可以在电脑上浏览WAP页面 很省流量的哈

《科学与工程数值计算算法（Visual C++版）》附盘的使用说明 =================================================================== 1、本书附赠的光盘包含了本书中全部的源代码。使用时只需将相应的目录拷贝到您的硬盘中。 注意 拷贝到硬盘上的源文件的属性如果成为只读的，在编译之前应该将它们的属性改为可读写的。 2、光盘各目录中的内容如下所示： 光盘目录 内容说明 \Source\ChapterN 第N章的所有示例工程源程序 \Source\Classes 本书所有算法类的源程序 \Source\Lib 集成本书所有算法的静态库文件 \Source\Dll 集成本书所有算法的动态库文件