带有 R Part II 的重量数据 [英] weight data with R Part II
问题描述
给定的是以下数据框:
structure(list(UH6401 = c(1, 1, 0, 0, 0, 1, 1, 0, 1, 0, 1, 1,
1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 0,
0, 0, 1, 1, 1, 1, 0, 1, 1, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 0, 0,
1, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 1, 0, 0, 1, 1, 0, 0,
1, 0, 1, 1), UH6402 = c(1, 1, 0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1,
0, 1, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0,
1, 0, 1, 1, 1, 0, 1, 0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0,
0, 1, 1, 1, 1, 0, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1,
1, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 1, 1, 0, 1,
0, 1, 1), UH6403 = c(1, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0,
1, 1, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 1, 1, 0,
1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1,
1, 1, 1, 1, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 0, 1, 1,
0, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0,
1, 1), UH6404 = c(0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 0, 1,
0, 1, 0, 1, 1, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1,
1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1,
1, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0,
0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1,
1), UH6409 = c(1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 0, 1, 0,
1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0,
0, 0, 1, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0,
1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0,
1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0
), UH6410 = c(1, 0, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 0,
1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 0, 1,
1, 0, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 1,
1, 0, 1, 1, 1, 0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 0,
0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 1, 1, 0, 1, 1, 1, 0, 1, 1, 1, 0
), UH6411 = c(0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 0,
1, 1, 1, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 1, 1, 1, 0, 1,
0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 0, 1,
1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 0, 1
), UH6412 = c(1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 0,
1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1
), UH6503 = c(1, 1, 1, 1, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 0,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 0, 0,
1, 1, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0,
1, 0, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1,
1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1
), UH66 = c(1, 1, 1, 1, 0, 1, 1, 1, 0, 1, 0, 1, 0, 1, 1, 0, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1),
UH68 = c(0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0,
0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0,
0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0), UH6501a = c(1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1), UH6405a = c(1,
0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0,
0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0,
0, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 0, 0, 1, 0, 0, 0, 1, 1,
1, 1, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0,
1, 0, 1, 1), UH6407a = c(1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0,
1, 1, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 1,
1, 0, 0, 1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 1, 0,
1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0,
0, 1, 1, 1, 0, 1, 1, 0, 0, 0, 0, 1, 1), weight = c(405.002592353822,
479.360356183825, 526.548105855472, 810.005184707644, 312.321528531308,
930.961115757095, 567.383058387095, 475.323944260643, 1226.91439266118,
517.086839792615, 1200.2669656949, 810.005184707644, 656.723784884795,
605.370463928298, 668.467435759576, 558.112457492436, 793.751055244424,
479.360356183825, 1226.91439266118, 1606.54816212786, 1657.48609449633,
300.803580980276, 605.370463928298, 1140.55078447979, 669.102760422943,
810.005184707644, 1657.48609449633, 305.569853371963, 2994.30343152033,
762.922030382216, 479.360356183825, 1147.36030437824, 668.467435759576,
517.086839792615, 479.360356183825, 399.141865860217, 656.723784884795,
913.364738988386, 312.321528531308, 569.10576379231, 775.630259688922,
1207.22952429547, 1053.09621171094, 1140.55078447979, 314.857225320909,
668.467435759576, 2416.57081451012, 573.680152189121, 396.875527622212,
605.370463928298, 1036.3159447043, 3088.62283807823, 569.10576379231,
1140.55078447979, 2416.57081451012, 1147.36030437824, 762.922030382216,
702.064141140629, 351.032070570315, 629.714450641817, 517.086839792615,
1996.20228768022, 828.743047248167, 475.323944260643, 920.185794495882,
793.751055244424, 796.08788273764, 1197.42559758065, 405.002592353822,
418.584343119327, 300.803580980276, 654.76828203733, 2740.09421696516,
351.032070570315, 1069.6202614693, 2094.91447516374, 399.141865860217,
654.76828203733, 1003.65414063441, 573.680152189121, 851.074587580641,
913.364738988386, 762.922030382216, 1034.17367958523, 573.680152189121,
479.360356183825, 3208.8607844079, 654.76828203733, 908.055695892447,
328.361892442398, 1036.3159447043, 702.064141140629, 613.457196330588,
601.607161960551, 567.383058387095, 479.360356183825, 306.261087672466,
920.185794495882, 654.76828203733, 828.743047248167)), .Names = c("UH6401",
"UH6402", "UH6403", "UH6404", "UH6409", "UH6410", "UH6411", "UH6412",
"UH6503", "UH66", "UH68", "UH6501a", "UH6405a", "UH6407a", "weight"
), row.names = c(NA, 100L), class = "data.frame")
在社会科学中，我们通常有一个权重变量，通过该变量的因子对案例(行)进行加权，以校正样本以适应例如按年龄段划分的人口.如果某行的权重变量为1.6"，则表示该行需要观察 1.6 次才能拟合基础总体.
In social science we often have a weight variable to weight a case (row) by the factor of that variable to correct the sample to fit e.g. the population by age classes. If the weight variable of a row is "1.6" it means that this row need do be observed 1.6 times to fit the basis population.
在 SPSS 中我会写
In SPSS I would write
WEIGHT BY weight.
该命令之后的所有程序都会相应地对数据进行加权.
and all procedures after that command will weight the data accordingly.
在 R 中，我可以使用命令 stabs 来做到这一点
In R I can do that with stabs with the command
xtabs(weight ~ UH6401, data=df)
但是如果我想进行 SVD 或 PCA 分析怎么办?这里没有像在 xtabs 中那样对数据进行加权的功能.
But what if I want to do a SVD or PCA analysis? Here there is no function to weight data like it is in xtabs.
所以问题是，有没有一种方法可以像在 SPSS 中那样对 R 中的数据进行加权?整数的点很容易，有了因子2"，我们就可以把线加倍，但是所有的小数因子是什么?
So the question is, is there a method to weight data in R like it is possible in SPSS? The point with whole numbers would be easy, with the factor "2" we would just double the line, but what is with all the factors that are decimal?
更新:
SVD 或 PCA 只是一个例子！采取任何其他统计程序.在社会科学中，样本从来都不是完美的，但要对样本数据进行统计分析，样本需要代表基本人口，而样本大多不代表.所以我们尝试用权重来解决这个问题，所以样本代表了基本人群！
The SVD or PCA was just an example! Take any other statistical procedure. In social science the samples are never perfect, but to do an statistical analysis with sample data, the sample needs to represent the basic population, but a sample mostly doesn't. So we try to fix that deficit with weights, so the sample represent the basic population!
推荐答案
您可能需要熟悉 R. Baron 的 RSiteSearch 和 Rseek 的搜索引擎:这是 Baron 网站上加权 PCA"的首批点击之一:
You probably need to get acquainted with the search engines for R. Baron's RSiteSearch and Rseek: This is one of the first hits on "weighted PCA" at Baron's site:
http://finzi.psych.upenn.edu/R/library/aroma.light/html/wpca.matrix.html
在对 Joris Meys 回应的评论中得到澄清后，答案通常是人们需要明确一个人需要样本权重与其他类型的权重.回归加权是通过 survey 包完成的.Lumley 的关于调查方法的书区分了三种类型的权重.(lm 函数中的权重"是方差权重，而不是样本权重.)
With the clarification in the comment to Joris Meys response, the answer is often that one needs to be clear that one is desires sample weights versus other types of weighting. Regression weighting is done with the survey package. Lumley's book on survey methods distinguishes among three types of weights. (The "weights" in the lm function are variance weights, NOT sample weights.)
注意:PCA 和因子分析(实验性)都包含在调查包中.所以也许 Dominick 的问题要求在回归方法中采用统一的加权方法有一个单一的答案".
Note: Both PCA and factor analysis (experimental) are included in the survey package. So maybe Dominick's question requestiong a unified approach to weighting in regression methods has a single "answer".
这篇关于带有 R Part II 的重量数据的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
