以不同的方式替换值

Question

我有如下数据：

| genome |time_1|time_2|time_3|time_4|
|genome_1|  28  |28    |NA    |NA    |
|genome_2|  2   |4     |5     |2     |
|genome_3|  12  |12    |12    |12    |
|genome_4|  2   |NA    |NA    |NA    |
|genome_5|  3   |3     |NA    |NA    |

我希望在每一行中用1替换相同的值(即至少出现两次)，用0替换一次的值。我想要的桌子应该是这样的：

| genome |time_1|time_2|time_3|time_4|
|genome_1|  1   |1     |NA    |NA    |
|genome_2|  1   |0     |0     |1     |
|genome_3|  1   |1     |1     |1     |
|genome_4|  0   |NA    |NA    |NA    |
|genome_5|  1   |1     |NA    |NA    |

有什么办法吗？非常感谢

Answer 1

在R基地，您将做：

id <-t(apply(df[-1], 1, function(x)duplicated(x, NA)|duplicated(x, NA, TRUE)))
cbind(df[1], NA^is.na(df[-1]) * id)


    genome time_1 time_2 time_3 time_4
1 genome_1      1      1     NA     NA
2 genome_2      1      0      0      1
3 genome_3      1      1      1      1
4 genome_4      0     NA     NA     NA
5 genome_5      1      1     NA     NA

你也可以：

 df %>%
  pivot_longer(starts_with("time"))%>%
  mutate(value = (duplicated(value, NA)|duplicated(value, NA, TRUE))* NA^is.na(value))%>%
  pivot_wider()

# A tibble: 5 x 5
  genome   time_1 time_2 time_3 time_4
  <chr>     <int>  <int>  <int>  <int>
1 genome_1      1      1     NA     NA
2 genome_2      1      0      0      1
3 genome_3      1      1      1      1
4 genome_4      1     NA     NA     NA
5 genome_5      1      1     NA     NA

Answer 2

从你似乎不关心联系的评论来看，下面是一种使用rowwise的潜在方法

library(dplyr)
data %>%
   rowwise() %>%
   mutate(across(time_1:time_4,
                 ~ case_when(is.na(.) ~ NA_real_,
                             . == tail(as.numeric(names(sort(table(c_across(time_1:time_4))))),1) ~ 1,
                             TRUE ~ 0)
           ))
## A tibble: 5 x 5
## Rowwise: 
#  genome   time_1 time_2 time_3 time_4
#  <chr>     <dbl>  <dbl>  <dbl>  <dbl>
#1 genome_1      1      1     NA     NA
#2 genome_2      1      0      0      1
#3 genome_3      1      1      1      1
#4 genome_4      1     NA     NA     NA
#5 genome_5      1      1     NA     NA

数据：

data <- structure(list(genome = c("genome_1", "genome_2", "genome_3", 
"genome_4", "genome_5"), time_1 = c(28L, 2L, 12L, 2L, 3L), time_2 = c(28L, 
4L, 12L, NA, 3L), time_3 = c(NA, 5L, 12L, NA, NA), time_4 = c(NA, 
2L, 12L, NA, NA)), class = "data.frame", row.names = c(NA, -5L
))

Answer 3

这似乎适用于我：

a= cbind(c(28,2,12,2,3),c(28,4,12,NA,3))
a= cbind(a,c(NA,5,12,NA,NA))
a=cbind(a, c(NA,2,12,NA,NA))
a = as.data.frame(a)

e = function(i){
  i = as.numeric(i)
  b= table(i)
  b= ifelse(b>1,1,0)
  b = as.data.frame(b)
 b = b[as.character(i),]
 return(b)
}
t(apply(a, 1, e))