r语言 - cor(data[， -1]， use = "complete.obs" ) 中的错误："x"必须是数字

Question

我对R完全陌生 - 老实说，真的不知道我在做什么。但我真的需要按照某人的建议用这些数据运行双变量/多变量回归，我被困住了。任何帮助将不胜感激。

rm(list=ls())  
setwd("C:/Users/Bogi/Documents/School/Honors Thesis/Voting and Economic Data")  
data<-read.csv("BOGDAN_DATA1.csv")  
head(data)
round(cor(data[,-1],use="complete.obs"),1)  
Error in cor(data[, -1], use = "complete.obs") : 'x' must be numeric

德普特

structure(list(REGION = structure(1:6, .Label = c("Altai Republic", 
"Altai Territory", "Amur Region", "Arkhangelsk Region", "Astrakhan region", 
"Belgorod region"), class = "factor"), PCT_CHANGE_VOTE = structure(c(2L, 
3L, 5L, 4L, 6L, 1L), .Label = c("-13%", "-16%", "-17%", "-25%", 
"-26%", "2%"), class = "factor"), PCT_CHANGE_GRP = structure(c(2L, 
1L, 4L, 3L, 3L, 4L), .Label = c("10%", "17%", "19%", "27%"), class =    "factor"), 
PCT_CHANGE_INFLATION = structure(c(1L, 2L, 1L, 3L, 3L, 2L
), .Label = c("-2%", "-3%", "-4%"), class = "factor"), PCT_CHANGE_UNEMP =   structure(c(5L, 
4L, 1L, 2L, 6L, 3L), .Label = c("-13%", "-14%", "-17%", "-3%", 
"5%", "7%"), class = "factor"), POVERTY = c(18.6, 22.6, 20.4, 
14.4, 14.2, 8.6), POP_AGE1 = c(25.8, 16.9, 18.5, 17.1, 17.8, 
15.2), POP_AGE2 = c(58.8, 59.6, 61.3, 60.4, 60.8, 60.3), 
POP_AGE3 = c(15.4, 23.5, 20.2, 22.5, 21.4, 24.5), POP_URBAN = c(28.7, 
55.2, 67, 76.2, 66.7, 66.4), POP_RURAL = c(71.3, 44.8, 33, 
23.8, 33.3, 33.6), COMPUTER = c(46.4, 54.5, 66.1, 74, 65.1, 
55.2), INTERNET = c(32.1, 41, 50.7, 66.5, 60, 50.7)), .Names = c("REGION", 
"PCT_CHANGE_VOTE", "PCT_CHANGE_GRP", "PCT_CHANGE_INFLATION", 
"PCT_CHANGE_UNEMP", "POVERTY", "POP_AGE1", "POP_AGE2", "POP_AGE3", 
"POP_URBAN", "POP_RURAL", "COMPUTER", "INTERNET"), row.names = c(NA, 
6L), class = "data.frame")

Answer 1

您可以循环列 2：5 （ lapply(data[2:5], ..) ），删除列 2：5 （ gsub('[%]',..) ） 中的%并将列转换为 numeric 列。gsub的输出将character类，通过as.numeric将其转换为numeric

data[2:5] <- lapply(data[2:5], function(x) 
                  as.numeric(gsub('[%]', '', x)))
Cor1 <- round(cor(data[-1],use="complete.obs"),1)  

或者您可以使用 shell 上的awk删除这些列中的%（假设, 作为分隔符）

awk 'BEGIN {OFS=FS=","} function SUB(F) {sub(/%/,"", $F)}{SUB(2);SUB(3);SUB(4);SUB(5)}1' Bogdan.csv > Bogdan2.csv

使用 read.csv 读取文件并运行cor

 dat1 <- read.csv('Bogdan2.csv')
 Cor2 <- round(cor(dat1[-1], use='complete.obs'), 1)
 identical(Cor1, Cor2)
 #[1] TRUE