如何在R中将数据从一列移动到另一列

Question

我正在尝试将数据从一列移动到另一列，因为基础表单填写不正确。

在表格中，它会询问有关家庭的信息，并询问每个成员的年龄 (AGE) 和性别 (SEX)，每个家庭最多允许 5 人。但是有些用户填写了人 1,3 和 4 的信息，但没有填写人 2 的任何信息，因为他们错误地填写了人 2，划掉了详细信息并将人 2 的详细信息填入了人 3 的框等。

数据是这样的（这个数据中ref 1和5是正确的，其他都是不正确的）

df <- data.frame(
  ref = c(1, 2, 3, 4, 5, 6),
  AGE1 = c(45, 36, 26, 47, 24, NA),
  AGE2 = c(NA, 24, NA, 13, 57, 28),
  AGE3 = c(NA, NA, 35, NA, NA, 26),
  AGE4 = c(NA, NA, 15, 11, NA, NA),
  AGE5 = c(NA, 15, NA, NA, NA, NA),
  SEX1 = c("M", "F", "M", "M", "M", NA),
  SEX2 = c(NA, "M", NA, "F", "F", "F"),
  SEX3 = c(NA, NA, "M", NA, NA, "M"),
  SEX4 = c(NA, NA, "F", "F", NA, NA),
  SEX5 = c(NA, "F", NA, NA, NA, NA)
)

这是当前表格的样子 （我已将 NA 替换为 - 以方便阅读）

ref	AGE1	AGE2	AGE3	AGE4	AGE5	SEX1	SEX2	SEX3	SEX4	SEX5
1	45	-	-	-	-	M	-	-	-	-
2	36	24	-	-	15	F	M	-	-	F
3	26	-	35	15	-	M	-	M	F	-
4	47	13	-	11	-	M	F	-	F	-
5	24	57	-	-	-	M	F	-	-	-
6	-	28	26	-	-	-	F	M	-	-

但我希望它看起来像这样

ref	AGE1	AGE2	AGE3	AGE4	AGE5	SEX1	SEX2	SEX3	SEX4	SEX5
1	45	-	-	-	-	M	-	-	-	-
2	36	24	15	-	-	F	M	F	-	-
3	26	35	15	-	-	M	M	F	-	-
4	47	13	11	-	-	M	F	F	-	-
5	24	57	-	-	-	M	F	-	-	-
6	28	26	-	-	-	F	M	-	-	-

有没有办法使用dplyr 来纠正这个问题？如果没有，R中是否有另一种方法来纠正数据

Answer 1

这是一种使用dplyr 和tidyr 的方法。该方法包括将数据转为较长格式，将NA 值排序到末尾，重新编号列名，然后再次转为宽格式。

library(dplyr)
library(tidyr)

df <-  data.frame(ref, AGE1, AGE2, AGE3, AGE4, AGE5,
                  SEX1, SEX2, SEX3, SEX4, SEX5)
df %>% 
  mutate(across(starts_with("AGE"), as.character)) %>% 
  pivot_longer(2:11) %>%
  separate(name, into = c("cat", "num"), 3) %>%
  arrange(is.na(value)) %>%
  group_by(ref, cat) %>%
  mutate(num = seq_along(value)) %>%
  ungroup() %>%
  arrange(cat) %>%
  unite(name, cat, num, sep = "") %>%
  pivot_wider(id_cols = ref) %>%
  mutate(across(starts_with("AGE"), as.numeric))


# A tibble: 6 x 11
    ref  AGE1  AGE2  AGE3  AGE4  AGE5 SEX1  SEX2  SEX3  SEX4  SEX5 
  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <chr> <chr> <chr> <chr> <chr>
1     1    45    NA    NA    NA    NA M     NA    NA    NA    NA   
2     2    36    24    15    NA    NA F     M     F     NA    NA   
3     3    26    35    15    NA    NA M     M     F     NA    NA   
4     4    47    13    11    NA    NA M     F     F     NA    NA   
5     5    24    57    NA    NA    NA M     F     NA    NA    NA   
6     6    28    26    NA    NA    NA F     M     NA    NA    NA  

Answer 2

试试下面的基本代码

u1 <- reshape(
  setNames(df, sub("(\\d)", ".\\1", names(df))),
  direction = "long",
  idvar = "ref",
  varying = -1
)

u2  <- reshape(
  transform(
    u1[with(u1, order(is.na(AGE), is.na(SEX))), ],
    time = ave(time, ref, FUN = seq_along)
  ),
  direction = "wide",
  idvar = "ref"
)

out <- u2[match(names(df),sub("\\.","",names(u2)))]

你会得到

> out
    ref AGE.1 AGE.2 AGE.3 AGE.4 AGE.5 SEX.1 SEX.2 SEX.3 SEX.4 SEX.5
1.1   1    45    NA    NA    NA    NA     M  <NA>  <NA>  <NA>  <NA>
2.1   2    36    24    15    NA    NA     F     M     F  <NA>  <NA>
3.1   3    26    35    15    NA    NA     M     M     F  <NA>  <NA>
4.1   4    47    13    11    NA    NA     M     F     F  <NA>  <NA>
5.1   5    24    57    NA    NA    NA     M     F  <NA>  <NA>  <NA>
6.2   6    28    26    NA    NA    NA     F     M  <NA>  <NA>  <NA>

数据

df <- data.frame(
  ref = c(1, 2, 3, 4, 5, 6),
  AGE1 = c(45, 36, 26, 47, 24, NA),
  AGE2 = c(NA, 24, NA, 13, 57, 28),
  AGE3 = c(NA, NA, 35, NA, NA, 26),
  AGE4 = c(NA, NA, 15, 11, NA, NA),
  AGE5 = c(NA, 15, NA, NA, NA, NA),
  SEX1 = c("M", "F", "M", "M", "M", NA),
  SEX2 = c(NA, "M", NA, "F", "F", "F"),
  SEX3 = c(NA, NA, "M", NA, NA, "M"),
  SEX4 = c(NA, NA, "F", "F", NA, NA),
  SEX5 = c(NA, "F", NA, NA, NA, NA)
)

Answer 3

这是一种使用dplyr 和tidyr 库的方法。

library(dplyr)
library(tidyr)

df %>%
  pivot_longer(cols = -ref, 
               names_to = c('.value', 'num'), 
               names_pattern = '([A-Z]+)(\\d+)') %>%
  arrange(ref, AGE, SEX) %>%
  group_by(ref) %>%
  mutate(num = row_number()) %>%
  ungroup %>%
  pivot_wider(names_from = num, values_from = c(AGE, SEX)) 

#    ref AGE_1 AGE_2 AGE_3 AGE_4 AGE_5 SEX_1 SEX_2 SEX_3 SEX_4 SEX_5
#  <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <chr> <chr> <chr> <chr> <chr>
#1     1    45    NA    NA    NA    NA M     NA    NA    NA    NA   
#2     2    15    24    36    NA    NA F     M     F     NA    NA   
#3     3    15    26    35    NA    NA F     M     M     NA    NA   
#4     4    11    13    47    NA    NA F     F     M     NA    NA   
#5     5    24    57    NA    NA    NA M     F     NA    NA    NA   
#6     6    26    28    NA    NA    NA M     F     NA    NA    NA    

Answer 4

这是使用包dedupewider的解决方案：

library(dedupewider)

df <- data.frame(
  ref = c(1, 2, 3, 4, 5, 6),
  AGE1 = c(45, 36, 26, 47, 24, NA),
  AGE2 = c(NA, 24, NA, 13, 57, 28),
  AGE3 = c(NA, NA, 35, NA, NA, 26),
  AGE4 = c(NA, NA, 15, 11, NA, NA),
  AGE5 = c(NA, 15, NA, NA, NA, NA),
  SEX1 = c("M", "F", "M", "M", "M", NA),
  SEX2 = c(NA, "M", NA, "F", "F", "F"),
  SEX3 = c(NA, NA, "M", NA, NA, "M"),
  SEX4 = c(NA, NA, "F", "F", NA, NA),
  SEX5 = c(NA, "F", NA, NA, NA, NA)
)

age_moved <- na_move(df, cols = names(df)[grepl("^AGE\\d$", names(df))]) # 'right' direction is by default

sex_moved <- na_move(age_moved, cols = names(df)[grepl("^SEX\\d$", names(df))])

sex_moved

#>   ref AGE1 AGE2 AGE3 AGE4 AGE5 SEX1 SEX2 SEX3 SEX4 SEX5
#> 1   1   45   NA   NA   NA   NA    M <NA> <NA>   NA   NA
#> 2   2   36   24   15   NA   NA    F    M    F   NA   NA
#> 3   3   26   35   15   NA   NA    M    M    F   NA   NA
#> 4   4   47   13   11   NA   NA    M    F    F   NA   NA
#> 5   5   24   57   NA   NA   NA    M    F <NA>   NA   NA
#> 6   6   28   26   NA   NA   NA    F    M <NA>   NA   NA