使用`lapply`跨数据框(R)中的所有列进行Winsoring

问题描述

我正在尝试使用library(DescTools)软件包中的lapply应用Winsorize()函数.我目前所拥有的是

I am trying to apply the Winsorize() function using lapply from the library(DescTools) package. What I currently have is;

data$col1 <- Winsorize(data$col1)

本质上将极端值替换为基于分位数的值，如下替换以下数据；

Which essentially replaces the extreme values with a value based on quantiles, replacing the below data as follows;

> data$col1
 [1]   -0.06775798   **-0.55213508**   -0.12338265
 [4]    0.04928349    **0.47524313**    0.04782829
 [7]   -0.05070639 **-112.67126382**    0.12657896
[10]   -0.12886632

> Winsorize(data$col1)
 [1] -0.06775798 **-0.37884540** -0.12338265  0.04928349
 [5]  **0.26038103**  0.04782829 -0.05070639 **-0.37884540**
 [9]  0.12657896 -0.12886632

我有一个for loop可以在data.frame col1，col2，col3，col4的所有列上执行此操作，但是，我知道lapply是更好的选择，所以我正在尝试将其合并到lapply函数中，但似乎无法使其正常工作.如果有人能指出我正确的方向，那将是非常感动的.

I have a for loop which can do this across all columns of the data.frame col1, col2, col3, col4, however, I know lapply is a better option so I am trying to incorporate it into an lapply function but cannot seem to get it working. If anybody can point me in the right direction it would be much apreciated.

数据；

data <- structure(list(EQ.TA = c(-0.0677579847115102, -0.552135083517749, 
-0.123382654164705, 0.0492834931482554, 0.475243125304193, 0.0478282913638668, 
-0.050706389027946, -112.671263815473, 0.126578956975704, -0.128866322940619
), NI.EQ = c(3.64670235329765, 1.66115713369585, 0.209424623633739, 
0.340430636358184, -0.248411254566261, -12.1709277350516, 1.06888235737433, 
0.0515582237132515, 0.177323118521857, 0.419879195374698), NI.TA = c(-0.24709320230217, 
-0.917183132749265, -0.0258393659113752, 0.0167776109344148, 
-0.118055740980805, -0.582114677880617, -0.0541991646381309, 
-5.80913022585296, 0.0224453753901758, -0.0541082879872031), 
    TL.TA = c(1.06775798471151, 1.55213508351775, 1.12338265416471, 
    0.950716506851745, 0.524756874695807, 0.952171708636133, 
    1.05070638902795, 113.671263815473, 0.873421043024296, 1.12886632294062
    )), .Names = c("EQ.TA", "NI.EQ", "NI.TA", "TL.TA"), row.names = c(NA, 
10L), class = "data.frame")

推荐答案

您可以在整个data.frame中lapply并重新分配它，例如:

You can lapply over the whole data.frame and reassign it like:

library(DescTools)
data[]<-lapply(data, Winsorize)

data
#          EQ.TA       NI.EQ       NI.TA      TL.TA
#1   -0.06775798  2.75320700 -0.24709320  1.0677580
#2   -0.55213508  1.66115713 -0.91718313  1.5521351
#3   -0.12338265  0.20942462 -0.02583937  1.1233827
#4    0.04928349  0.34043064  0.01677761  0.9507165
#5    0.31834425 -0.24841125 -0.11805574  0.6816558
#6    0.04782829 -6.80579532 -0.58211468  0.9521717
#7   -0.05070639  1.06888236 -0.05419916  1.0507064
#8  -62.21765589  0.05155822 -3.60775403 63.2176559
#9    0.12657896  0.17732312  0.01989488  0.8734210
#10  -0.12886632  0.41987920 -0.05410829  1.1288663

这篇关于使用`lapply`跨数据框(R)中的所有列进行Winsoring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！