RSM

说明：  Optics聚类算法 OPTICS没有显示地产生一个数据集合簇，它为自动和交互地聚类分析计算一个簇次序。这个次序代表了数据基于密度地聚类结构。它包含地信息，等同于从一个宽广地参数设置范围所获得的基于密度的聚类(Optics do not show clustering algorithm OPTICS to produce a collection of data clusters, it is automatically and interactively computing cluster analysis a cluster order. This order represents the data to cluster based on the density structure. It contains in information from a broadly equivalent range of parameters obtained by density-based clustering)

这是一个最最简单计算器，提供给大家参考一下(dadasddasdasd fafdsafasfafad)

程序是关于计算两相黎曼问题的一个求解器。可以直接设置。(The code is a solver about the two-phase Riemann problem. the user can use it directly.)

fluent软件的UDF程序，是焊接过程的热源，热源类型为双椭球空间热源。(fluent udf heat source)

wolfe程序模块的C++算法，可以重复使用,对于解决一维搜索是很好的..(C++ algorithm wolfe program modules can be reused for solving one-dimensional search is good ..)

实现1到1000位的大数运算（包括加减乘除和浮点数运算），运行环境是vs2005(1-1000, large number operations (addition, subtraction and floating-point), the operating environment is vs2005)

偏最小二乘法中的交叉有效性计算,看看，可能有用(PLS cross-validity of the calculation, see, may be useful)

说明：  LDL分解。作为矩阵方程数值解法最基础的矩阵分解算法，LDL算法可以高效分解对称矩阵。(LDL decomposition. Numerical Solution of matrix equation as the most basic matrix factorization algorithm, LDL decomposition algorithm can be efficient symmetric matrix.)

计算按频率抽取的基4快速傅里叶变换算法，与基2FFT相比，基4FFT的乘法量约减少25 。(Calculating frequency taken by fast Fourier transform algorithm base 4, compared with the base 2FFT the base 4FFT multiplication about 25 less.)

之前找不到合适精度的log2x函数的定点化源码，经过实现并验证后，发上来大家一起学习(fixed version for log2x)