找回密码
 注册
查看: 705|回复: 0

R语言 qpgraph包 qpHTF()函数中文帮助文档(中英文对照)

[复制链接]
发表于 2012-2-26 11:40:15 | 显示全部楼层 |阅读模式
qpHTF(qpgraph)
qpHTF()所属R语言包:qpgraph

                                         Hastie Tibshirani Friedman algorithm
                                         哈斯蒂Tibshirani弗里德曼算法

                                         译者:生物统计家园网 机器人LoveR

描述----------Description----------

Performs maximum likelihood estimation of a covariance matrix given the independence constraints from an input undirected graph.
执行输入无向图的独立约束的协方差矩阵的最大似然估计。


用法----------Usage----------


qpHTF(S, g, tol = 0.001, verbose = FALSE, R.code.only = FALSE)



参数----------Arguments----------

参数:S
input matrix, in the context of this package, the sample covariance matrix.
输入矩阵,在这个包的情况下,样本协方差矩阵。


参数:g
input undirected graph.
输入无向图。


参数:tol
tolerance under which the iterative algorithm stops.
宽容下的迭代算法停止。


参数:verbose
show progress on calculations.
显示在计算方面取得的进展。


参数:R.code.only
logical; if FALSE then the faster C implementation is used (default); if TRUE then only R code is executed.
逻辑;如果为FALSE则更快的C实现使用(默认);如果TRUE,那么只有R代码被执行。


Details

详情----------Details----------

This is an alternative to the Iterative Proportional Fitting (IPF) algorithm (see, Whittaker, 1990, pp. 182-185 and qpIPF) which also adjusts the input matrix to the independence constraints in the input undirected graph. However, differently to the IPF, it works by going through each of the vertices fitting the marginal distribution over the corresponding vertex boundary. It stops when the adjusted matrix at the current iteration differs from the matrix at the previous iteration in less or equal than a given tolerance value. This algorithm is described by Hastie, Tibshirani and Friedman (2009, pg. 634), hence we name it here HTF, and it has the advantage over the IPF that it does not require the list of maximal cliques of the graph which may be exponentially large. In contrast, it requires that the maximum boundary size of the graph is below the number of samples where the input sample covariance matrix S was estimated. For the purpose of exploring qp-graphs that meet such a requirement, one can use the function qpBoundary.
这是一个替代的比例拟合迭代算法(IPF)的(见,惠特克,1990年,第182-185和qpIPF)也调整输入矩阵输入无向图的独立约束。然而,不同的森林小组,它的工作原理是通过拟合了相应的顶点边界的边缘分布的每个顶点。它停止时,在当前迭代的调整矩阵矩阵不同,在上一次迭代小于或等于给定的公差值。该算法描述黑斯蒂,Tibshirani和弗里德曼(2009年,634页),因此,我们在这里将其命名为导热油,它具有在IPF的优势在于,它不需要图可能是指数的最大派系的列表大。相反,它需要的图的最大边界的大小是在下面的输入样本协方差矩阵S估计的样本数。为探索QP图,以满足这种要求的目的,可以使用的功能qpBoundary。


值----------Value----------

The input matrix adjusted to the constraints imposed by the input undirected graph, i.e., a maximum likelihood estimate of the sample covariance matrix that includes the independence constraints encoded in the undirected graph.
调整输入矩阵输入无向图,即最大似然估计的样本协方差矩阵,其中包括在无向图的独立编码约束的限制。


注意----------Note----------

Thanks to Giovanni Marchetti for bringing us our attention to this algorithm and sharing an early version of its implementation on the R package ggm.
感谢为我们带来了我们的注意,该算法和共享其实施的早期版本的R包ggm乔瓦尼·马尔凯蒂。


作者(S)----------Author(s)----------


R. Castelo



参考文献----------References----------

Gaussian graphical model search from microarray data with p larger than n. J. Mach. Learn. Res., 7:2621-2650, 2006.

Wiley, 1990.

参见----------See Also----------

qpBoundary qpIPF qpPAC
qpBoundaryqpIPFqpPAC


举例----------Examples----------


require(graph)
require(mvtnorm)

nVar <- 50  ## number of variables[#变量]
nObs <- 100 ## number of observations to simulate[#号观测到模拟]

set.seed(123)

g <- randomEGraph(as.character(1:nVar), p=0.15)

Sigma <- qpG2Sigma(g, rho=0.5)
X <- rmvnorm(nObs, sigma=as.matrix(Sigma))

## MLE of the sample covariance matrix[#MLE的样本协方差矩阵]
S <- cov(X)

## more efficient MLE of the sample covariance matrix using HTF[#MLE的更有效的使用导热油的样本协方差矩阵]
S_htf <- qpHTF(S, g)

## get the adjacency matrix and put the diagonal to one[#得到的邻接矩阵和一个对角线]
A <- as(g, "matrix")
diag(A) <- 1

## entries in S and S_htf for present edges in g should coincide[在S和S_htf#项目目前在G边缘应一致]
max(abs(S_htf[A==1] - S[A==1]))

## entries in the inverse of S_htf for missing edges in g should be zero[#G的边失踪中的S_htf的逆的条目应该为零]
max(solve(S_htf)[A==0])

转载请注明:出自 生物统计家园网(http://www.biostatistic.net)。


注:
注1:为了方便大家学习,本文档为生物统计家园网机器人LoveR翻译而成,仅供个人R语言学习参考使用,生物统计家园保留版权。
注2:由于是机器人自动翻译,难免有不准确之处,使用时仔细对照中、英文内容进行反复理解,可以帮助R语言的学习。
注3:如遇到不准确之处,请在本贴的后面进行回帖,我们会逐渐进行修订。
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 注册

本版积分规则

手机版|小黑屋|生物统计家园 网站价格

GMT+8, 2025-1-31 22:37 , Processed in 0.028731 second(s), 15 queries .

Powered by Discuz! X3.5

© 2001-2024 Discuz! Team.

快速回复 返回顶部 返回列表