使用LOOP将单元格中的缺失值替换为n-1以上单元格中的值

问题描述

我有一个包含数千行的数据文件，其中有一些空白，我希望用一个值来填充。 我需要将空单元格替换为其上方的值。 让您更容易了解我的数据，这是一个示例

Variable <- c("AGE","","SEX","SEGMENT","")    
Value <- c(1,2,3,4,1,5)
Description <- c("18-24","25-34","35-44","45+","Female","Male","A","B","C","D","E")
df <- data.frame(Variable,Value,Description)

> df
   Variable Value Description
1       AGE     1       18-24
2               2       25-34
3               3       35-44
4               4         45+
5       SEX     1      Female
6               2        Male
7   SEGMENT     1           A
8               2           B
9               3           C
10              4           D
11              5           E

您可以在第一列上方看到空白。我需要将这些空单元格替换为上面的相关值，以便新变量在数据框中看起来像这样

> df
   Variable Value Description Variable_NEW
1       AGE     1       18-24               AGE
2               2       25-34               AGE
3               3       35-44               AGE
4               4         45+               AGE
5       SEX     1      Female               SEX
6               2        Male               SEX
7   SEGMENT     1           A           SEGMENT
8               2           B           SEGMENT
9               3           C           SEGMENT
10              4           D           SEGMENT
11              5           E           SEGMENT

大声思考。我假设要实现此目标，我将需要使用循环创建一个新变量，然后使用类似这样的逻辑

    IF Variable[n]="" THEN Variable_New[n] = Variable[n-1],ELSE Variable_New[n] = Variable[n]

我熟悉循环，但是不知道如何在具有滞后/ n-1函数的R中编写这种东西。可能有很多方法可以完成此操作，但是最好使用循环。任何帮助将不胜感激。谢谢

解决方法

这里有一个循环方法：

#Data
Variable <- c("AGE","","SEX","SEGMENT","")    
Value <- c(1,2,3,4,1,5)
Description <- c("18-24","25-34","35-44","45+","Female","Male","A","B","C","D","E")
df <- data.frame(Variable,Value,Description,stringsAsFactors = F)
#Create new column
df$NewVar <- df$Variable
#Loop
for(i in 2:dim(df)[1])
{
  df$NewVar[i] <- ifelse(df$NewVar[i]=="",df$NewVar[i-1],df$NewVar[i])
}

输出：

   Variable Value Description  NewVar
1       AGE     1       18-24     AGE
2               2       25-34     AGE
3               3       35-44     AGE
4               4         45+     AGE
5       SEX     1      Female     SEX
6               2        Male     SEX
7   SEGMENT     1           A SEGMENT
8               2           B SEGMENT
9               3           C SEGMENT
10              4           D SEGMENT
11              5           E SEGMENT

,

您无需编写循环，内置的函数可以帮助您完成此任务。

您可以使用replace NA空白值并使用fill：

library(dplyr)

df %>%
  mutate(Variable_NEW = replace(Variable,Variable == "",NA)) %>%
  tidyr::fill(Variable_NEW)

#   Variable Value Description Variable_NEW
#1       AGE     1       18-24          AGE
#2               2       25-34          AGE
#3               3       35-44          AGE
#4               4         45+          AGE
#5       SEX     1      Female          SEX
#6               2        Male          SEX
#7   SEGMENT     1           A      SEGMENT
#8               2           B      SEGMENT
#9               3           C      SEGMENT
#10              4           D      SEGMENT
#11              5           E      SEGMENT

,

您可以使用循环编写自己的函数，也可以使用na.locf包中的zoo函数来填写缺少的NA值。示例：

fillin <- function(x) {
  for (i in 2:length(x)) {
    if (x[i] %in% c(NA,"")) {
      x[i] <- x[i - 1]
    }
  }
  x
}

Variable <- c("AGE",Description)

df$Variable_fillin <- fillin(df$Variable)

library(zoo)
df$Variable[df$Variable == ""] <- NA
df$Variable_nalocf <- na.locf(df$Variable)

df
#>    Variable Value Description Variable_fillin Variable_nalocf
#> 1       AGE     1       18-24             AGE             AGE
#> 2      <NA>     2       25-34             AGE             AGE
#> 3      <NA>     3       35-44             AGE             AGE
#> 4      <NA>     4         45+             AGE             AGE
#> 5       SEX     1      Female             SEX             SEX
#> 6      <NA>     2        Male             SEX             SEX
#> 7   SEGMENT     1           A         SEGMENT         SEGMENT
#> 8      <NA>     2           B         SEGMENT         SEGMENT
#> 9      <NA>     3           C         SEGMENT         SEGMENT
#> 10     <NA>     4           D         SEGMENT         SEGMENT
#> 11     <NA>     5           E         SEGMENT         SEGMENT

,

这用缺少的字符替换了“”，然后修复了名为Variable的变量：

df %>% 
  dplyr::mutate_all(list(~na_if(.,""))) %>% 
  tidyr::fill(Variable,.direction = "down")

,

使用data.table和for循环：

library(data.table)
DT <- as.data.table(df)

DT[,Variable_new := Variable[1]]

for (i in 2:nrow(DT)) {
  DT[i,Variable_new := fifelse(DT[i,Variable] == '',DT[i-1,Variable_new],DT[i,Variable])]
}

> DT
    Variable Value Description Variable_new
 1:      AGE     1       18-24          AGE
 2:              2       25-34          AGE
 3:              3       35-44          AGE
 4:              4         45+          AGE
 5:      SEX     1      Female          SEX
 6:              2        Male          SEX
 7:  SEGMENT     1           A      SEGMENT
 8:              2           B      SEGMENT
 9:              3           C      SEGMENT
10:              4           D      SEGMENT
11:              5           E      SEGMENT