如果我们的数据帧中缺少数据,那么如果我们有足够的有关丢失信息的案件特征的信息,则可以替换其中一些信息。但是,如果该信息不可用,并且我们找不到任何合适的方法来替换缺失值,则可以对具有缺失值的列使用complete.cases函数。
请看以下数据帧:
> set.seed(19991) > x1<-sample(c(NA,rnorm(5,2,1)),20,replace=TRUE) > x2<-sample(c(NA,rnorm(5,40,0.87)),20,replace=TRUE) > x3<-sample(c(NA,rnorm(5,1,0.015)),20,replace=TRUE) > x4<-sample(c(NA,rnorm(10,5,1.27)),20,replace=TRUE) > x5<-sample(c(NA,rnorm(8,1,0.20)),20,replace=TRUE) > df1<-data.frame(x1,x2,x3,x4,x5) > df1
输出结果
x1 x2 x3 x4 x5 1 0.8287962 39.74094 0.9983586 6.338327 0.8692225 2 1.3167347 NA NA 4.133738 0.8692225 3 3.9911408 38.84212 1.0047761 5.825111 0.8423061 4 0.6426335 39.74094 1.0047761 5.177329 NA 5 1.3167347 NA 0.9963252 5.073915 0.8423061 6 0.8287962 38.84212 0.9963252 5.154073 1.0566156 7 NA 40.36844 0.9927987 NA 0.8423061 8 0.1952913 40.36844 1.0047761 6.338327 NA 9 3.9911408 NA 1.0366262 5.154073 1.1936387 10 0.6426335 39.77818 0.9927987 5.177329 0.8557775 11 NA NA 1.0047761 7.216787 0.9506370 12 NA 38.84212 0.9983586 NA 0.8423061 13 1.3167347 39.77818 0.9963252 5.825111 0.8557775 14 0.8287962 39.77818 1.0366262 5.177329 NA 15 0.1952913 NA 0.9927987 5.073915 0.8692225 16 0.1952913 38.84212 1.0366262 5.154073 0.8286973 17 0.1952913 38.84212 1.0366262 NA 0.9506370 18 1.3167347 40.36844 0.9983586 NA 1.0566156 19 0.1952913 39.80231 NA 5.073915 NA 20 NA NA 0.9983586 5.073915 0.8557775
删除df1行,其中第3至5列包含NA:
> df1[complete.cases(df1[3:5]),]
输出结果
x1 x2 x3 x4 x5 1 0.8287962 39.74094 0.9983586 6.338327 0.8692225 3 3.9911408 38.84212 1.0047761 5.825111 0.8423061 5 1.3167347 NA 0.9963252 5.073915 0.8423061 6 0.8287962 38.84212 0.9963252 5.154073 1.0566156 9 3.9911408 NA 1.0366262 5.154073 1.1936387 10 0.6426335 39.77818 0.9927987 5.177329 0.8557775 11 NA NA 1.0047761 7.216787 0.9506370 13 1.3167347 39.77818 0.9963252 5.825111 0.8557775 15 0.1952913 NA 0.9927987 5.073915 0.8692225 16 0.1952913 38.84212 1.0366262 5.154073 0.8286973 20 NA NA 0.9983586 5.073915 0.8557775
删除df1行,其中第1到3列包含NA:
> df1[complete.cases(df1[1:3]),]
输出结果
x1 x2 x3 x4 x5 1 0.8287962 39.74094 0.9983586 6.338327 0.8692225 3 3.9911408 38.84212 1.0047761 5.825111 0.8423061 4 0.6426335 39.74094 1.0047761 5.177329 NA 6 0.8287962 38.84212 0.9963252 5.154073 1.0566156 8 0.1952913 40.36844 1.0047761 6.338327 NA 10 0.6426335 39.77818 0.9927987 5.177329 0.8557775 13 1.3167347 39.77818 0.9963252 5.825111 0.8557775 14 0.8287962 39.77818 1.0366262 5.177329 NA 16 0.1952913 38.84212 1.0366262 5.154073 0.8286973 17 0.1952913 38.84212 1.0366262 NA 0.9506370 18 1.3167347 40.36844 0.9983586 NA 1.0566156
删除df1中第2至4列包含NA的行:
> df1[complete.cases(df1[2:4]),]
输出结果
x1 x2 x3 x4 x5 1 0.8287962 39.74094 0.9983586 6.338327 0.8692225 3 3.9911408 38.84212 1.0047761 5.825111 0.8423061 4 0.6426335 39.74094 1.0047761 5.177329 NA 6 0.8287962 38.84212 0.9963252 5.154073 1.0566156 8 0.1952913 40.36844 1.0047761 6.338327 NA 10 0.6426335 39.77818 0.9927987 5.177329 0.8557775 13 1.3167347 39.77818 0.9963252 5.825111 0.8557775 14 0.8287962 39.77818 1.0366262 5.177329 NA 16 0.1952913 38.84212 1.0366262 5.154073 0.8286973
让我们看另一个例子:
> y1<-sample(c(NA,rpois(5,2)),20,replace=TRUE) > y2<-sample(c(NA,rpois(5,5)),20,replace=TRUE) > y3<-sample(c(NA,rpois(5,1)),20,replace=TRUE) > y4<-sample(c(NA,rpois(5,2)),20,replace=TRUE) > df2<-data.frame(y1,y2,y3,y4) > df2
输出结果
y1 y2 y3 y4 1 0 2 0 NA 2 6 NA NA NA 3 0 9 1 1 4 6 4 NA 1 5 2 2 0 2 6 2 9 NA NA 7 6 2 0 1 8 2 4 1 NA 9 2 2 1 1 10 6 4 1 2 11 2 2 0 NA 12 6 2 3 1 13 0 4 1 1 14 2 4 1 0 15 2 9 0 1 16 2 2 1 1 17 2 9 NA 1 18 2 9 0 1 19 2 9 1 0 20 NA 2 3 1
> df2[complete.cases(df2[1:3]),]
输出结果
y1 y2 y3 y4 1 0 2 0 NA 3 0 9 1 1 5 2 2 0 2 7 6 2 0 1 8 2 4 1 NA 9 2 2 1 1 10 6 4 1 2 11 2 2 0 NA 12 6 2 3 1 13 0 4 1 1 14 2 4 1 0 15 2 9 0 1 16 2 2 1 1 18 2 9 0 1 19 2 9 1 0
> df2[complete.cases(df2[2:4]),]
输出结果
y1 y2 y3 y4 3 0 9 1 1 5 2 2 0 2 7 6 2 0 1 9 2 2 1 1 10 6 4 1 2 12 6 2 3 1 13 0 4 1 1 14 2 4 1 0 15 2 9 0 1 16 2 2 1 1 18 2 9 0 1 19 2 9 1 0 20 NA 2 3 1
> df2[complete.cases(df2[c(1,3)]),]
输出结果
y1 y2 y3 y4 1 0 2 0 NA 3 0 9 1 1 5 2 2 0 2 7 6 2 0 1 8 2 4 1 NA 9 2 2 1 1 10 6 4 1 2 11 2 2 0 NA 12 6 2 3 1 13 0 4 1 1 14 2 4 1 0 15 2 9 0 1 16 2 2 1 1 18 2 9 0 1 19 2 9 1 0