L1median(robustX)
L1median()所属R语言包:robustX
Compute the Multivariate L1-Median
L1-中位数计算多元
译者:生物统计家园网 机器人LoveR
描述----------Description----------
Compute the multivariate L1-median m, i.e., the minimizer of
计算的多元L1中位数m,即最小化
where || u || = sqrt(sum(j = 1:p; u[j]^2)).
|| u || = sqrt(sum(j = 1:p; u[j]^2))。
As a convex problem, there's always a global minimizer, computable not by a closed formula but rather an iterative search. As the (partial) first derivatives of the objective function is undefined the data points, the minimization is not entirely trivial.
作为一个凸优化问题,总有一个全局极小,可计算的,而不是由一个封闭的公式,而是一个迭代搜索。由于(局部的)的目标函数的一阶导数是未定义的数据点,最小化是不完全微不足道。
用法----------Usage----------
L1median(X, m.init = apply(X, 2, median), weights = NULL,
method = c("nlm", "HoCrJo", "VardiZhang", optimMethods, nlminbMethods),
pscale = apply(abs(centr(X, m.init)), 2, mean, trim = 0.40),
tol = 1e-08, maxit = 200, trace = FALSE,
zero.tol = 1e-15, ...)
参数----------Arguments----------
参数:X
numeric matrix of dimension n x p, say.
数字matrix尺寸n x p“说。
参数:m.init
starting value for m; typically and by default the coordinatewise median.
初始值m;通常默认情况下,coordinatewise中位数。
参数:weights
optional numeric vector of non-negative weights; currently only implemented for method "VardiZhang".
可选的数字向量非负权重,目前只执行为方法"VardiZhang"。
参数:method
character string specifying the computational method, i.e., the algorithm to be used (can be abbreviated).
字符串指定的计算方法,也就是说,该算法使用(可以缩写)。
参数:pscale
numeric p-vector of positive numbers, the coordinate-wise scale (typical size of delta(m[j])), where m is the problem's solution.
数值对向量的正数,坐标明智的规模(典型尺寸delta(m[j])),其中m的问题的解决方案。
参数:tol
positive number specifying the (relative) convergence tolerance.
正数指定的(相对)收敛公差。
参数:maxit
positive integer specifying the maximal number of iterations (before the iterations are stopped prematurely if necessary).
正整数,指定最大迭代次数(前提前终止的迭代如果必要的话)。
参数:trace
an integer specifying the tracing level of the iterations; 0 does no tracing
一个整数,指定跟踪级别的迭代0并没有追踪
参数:zero.tol
for method "VardiZhang", a small positive number specifying the tolerance for determining that the iteration is "exactly" at a data point (which is singularity).
方法"VardiZhang",一个小的正数,指定的公差确定,迭代是“准确”数据点(这是奇点)。
参数:...
optional arguments to nlm() or the control (list) arguments of optim(), or nlminb(), respectively.
可选参数nlm()或control(名单)参数optim()或nlminb(),。
Details
详细信息----------Details----------
Currently, we have to refer to the “References” below.
目前,我们有请参见下面的“参考”。
值----------Value----------
currently the result depends strongly on the method used.
目前的结果在很大程度上取决于method使用。
FIXME. This will change considerably.
FIXME。这将发生较大的变化。
(作者)----------Author(s)----------
Martin Maechler. Method <code>"HoCrJo"</code> is mostly based on Kristel
Joossens' function, implementing Hossjer and Croux (1995).
参考文献----------References----------
Generalizing Univariate Signed Rank Statistics for Testing and Estimating a Multivariate Location Parameter. Non-parametric Statistics 4, 293–308.
The multivariate <code>L_1</code>-median and associated data depth. Proc. National Academy of Science 97(4), 1423–1426.
参见----------See Also----------
median, covMcd
median,covMcd
实例----------Examples----------
data(stackloss)
L1median(stackloss)
L1median(stackloss, method = "HoCrJo")
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。
|