makegrid(wavethresh)
makegrid()所属R语言包：wavethresh
Interpolate data to a grid.
内插数据到电网。
译者：生物统计家园网 机器人LoveR
描述----------Description----------
This function takes a set of univariate (x,y) data with x arbitrary in (0,1) and linearly interpolates (x,y) to an equally spaced dyadic grid.
此功能需要一组单变量（X，Y）与x的任意数据（0,1）和线性内插（X，Y）的等距二元电网。
用法----------Usage----------
makegrid(t, y, gridn = 2^(floor(log(length(t)-1,2)) + 1))
参数----------Arguments----------
参数：t
A vector of x data. Each of the entries of x must lie between 0 and 1.
一个矢量的x数据。每个条目x必须位于0和1之间。
参数：y
A vector of y data. Each entry of y corresponds to the same-positioned entry in x and soy must be of the same length as x.
一个矢量的y数据。 y的每个条目对应于相同的位置的条目在x等y必须是相同的长度x。
参数：gridn
The number of grid points in the dyadic grid that the (x,y) gets interpolated to. By default this is the next power of two larger than the length of x.
在二进网格的网格点的数目，得到的（x，y）的内插。默认情况下，这是未来的力量大于x的长度。
Details
详细信息----------Details----------
One method for performing wavelet regression on data that is not equally spaced nor of power of two length is that described in Kovac, (1997) and Kovac and Silverman, (2000).
Kovac的，（1997）和科瓦奇和Silverman，（2000年）中所描述的数据是不等距隔开或两个长度的功率进行小波回归的方法之一。
The Kovac-Silverman algorithm linearly interpolates arbitrarily spaced (x,y) data to a dyadic grid and applies wavelet shrinkage to the interpolated data. However, if one assumes that the original data obeys a signal+noise model with iid data the interpolated data will be correlated due to the interpolation. This fact needs to be taken into account after taking the DWT and before thresholding one realizes that each coefficient has its own variance. The Kovac-Silverman algorithm computes this variance efficiently using knowledge of the interpolation scheme.
科瓦奇西尔弗曼算法线性插值任意间隔开的（的x，y）的数据到一个二进电网和应用小波收缩的内插数据。然而，如果一个假定原始数据服从一个信号+噪声模型与iid数据的内插数据将由于内插相关。这其实需要考虑的DWT后，阈值1来实现之前，每个系数都有自己的方差。该的科瓦奇 - 西尔弗曼算法计算方差有效地利用知识的插值方法。
值----------Value----------
An object of class griddata.
对象的类griddata。
RELEASE----------RELEASE----------
Version 3.9.6 Copyright Arne Kovac 1997
版本3.9.6版权所有1997年阿恩·科瓦奇
（作者）----------Author(s)----------
Arne Kovac
参见----------See Also----------
accessc, irregwd, newsure, plot.irregwd, threshold.irregwd,
accessc，irregwd，newsure，plot.irregwd，threshold.irregwd，
实例----------Examples----------
#[]
# Generate some values in (0,1), then sort them (for plotting)[生成（0,1）中的一些值，然后对其进行排序（用于绘图）]
#[]
tt <- sort(runif(100))
#[]
# Now evaluate the \code{\link{doppler}} function and add[我们评估了\ code {的\的链接{多普勒}}功能，并添加]
# some noise.[一些噪音。]
#[]
yy <- doppler(tt) + rnorm(100, 0, 0.15)
#[]
# Now make the grid with this data[现在做的这个数据网格]
#[]
yygrid <- makegrid(t=tt, y=yy)
#[]
# Jolly good. Now let's take the wavelet transform of this gridded data.[乔利好。现在，让我们来此栅格数据的小波变换的。]
# Note that we have to use the \code{\link{irregwd}} function[请注意，我们使用了\ code {\ {irregwd}}]
# of the gridded data as it computes the variances of the coefficients[的格点数据，因为它计算的方差系数]
# as well as the coefficients themselves.[以及系数本身。]
#[]
yyirregwd <- irregwd(yygrid)
#[]
# You might want to plot the coefficients[您可能要绘制的系数]
#[]
# If you want to see the actual coefficients you have to first convert[如果你想看到实际的系数，你必须先转换]
# the class of the yyirregwd object to a wd object and then use[类的yyirregwd对象到WD对象，然后使用]
# \code{\link{plot.wd}} like this[\代码{\的链接{plot.wd}}这样的]
#[]
yyirregwd2 <- yyirregwd
class(yyirregwd2) <- "wd"
## Not run: plot(yyirregwd2)[＃不运行：图（yyirregwd2）]
#[]
# If you want to see the variance factors (essentially the coefficient[如果你想看到的差异因素（主要是系数]
# variances divided by the overall variance). Then just use[方差除以总体方差）。然后，只需使用]
# \code{\link{plot.irregwd}}[\代码{\的链接{plot.irregwd}}]
#[]
## Not run: plot(yyirregwd)[＃不运行：图（yyirregwd）]
#[]
# Ok. So you've seen the coefficients. Now let's do some thresholding.[好吧。所以，你见过系数。现在，让我们做一些阈值。]
#[]
yy.thresh.sure <- threshold(yyirregwd, policy="sure", type="soft", dev=madmad)
#[]
# And now do the reconstruct[现在做的重建]
#[]
yy.wr <- wr(yy.thresh.sure)
#[]
# And you can even plot the answer on the new grid![你甚至可以在新的网格上绘制的答案！]
#[]
## Not run: plot(yygrid$gridt, yy.wr, type="l")[＃未运行图（yygrid gridt，yy.wr，类型为“L”）]
#[]
# And superimpose the original data![叠加的原始数据！]
#[]
## Not run: points(tt, yy)[＃不执行点（TT，YY）]
#[]
# This is sort of \code{Doppler} like![这是一种\的代码{多普勒}一样！]
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。
|
发表于 2012-10-1 17:25:42
