Question

当存在空白/缺失值时，如何才能使下面的数据框行唯一地依赖于第二列？

> head(interproscan)
                V1        V14
1 sp0000001-mRNA-1           
2 sp0000001-mRNA-1           
3 sp0000001-mRNA-1           
4 sp0000005-mRNA-1 GO:0003723
5 sp0000006-mRNA-1 GO:0016021
6 sp0000006-mRNA-1 GO:0016021


> head(unique(interproscan[ , 1:2] ))
                 V1                              V14
1  sp0000001-mRNA-1                                 
4  sp0000005-mRNA-1                       GO:0003723
5  sp0000006-mRNA-1                       GO:0016021
7  sp0000006-mRNA-2                       GO:0016021
9  sp0000006-mRNA-3                       GO:0016021

目标是：

                 V1                              V14
1  sp0000001-mRNA-1                                 
4  sp0000005-mRNA-1                       GO:0003723
5  sp0000006-mRNA-1                       GO:0016021

提前谢谢

Answer 1

您需要修改V1按照您的意图进行分组。我使用gsub丢弃最后-number个后缀。

library(dplyr)
ans <- df %>%
         group_by(gsub("-\\d","",V1), V14) %>%   # now it groups the way you want
         arrange(V1) %>%   # unnecessary for your toy example but just in case for your full data
         slice(1) %>%     # select top row-entry
         ungroup() %>%
         select(-4)     # discard intermediate grouping variable

输出

# A tibble: 3 x 3
     id               V1        V14
  <int>            <chr>      <chr>
1     1 sp0000001-mRNA-1           
2     4 sp0000005-mRNA-1 GO:0003723
3     5 sp0000006-mRNA-1 GO:0016021

数据

df <- structure(list(id = c(1L, 2L, 3L, 4L, 5L, 6L, 7L, 9L), V1 = c("sp0000001-mRNA-1", 
"sp0000001-mRNA-1", "sp0000001-mRNA-1", "sp0000005-mRNA-1", "sp0000006-mRNA-1", 
"sp0000006-mRNA-1", "sp0000006-mRNA-2", "sp0000006-mRNA-3"), 
    V14 = c("", "", "", "GO:0003723", "GO:0016021", "GO:0016021", 
    "GO:0016021", "GO:0016021")), class = "data.frame", .Names = c("id", 
"V1", "V14"), row.names = c(NA, -8L))


  id               V1        V14
1  1 sp0000001-mRNA-1           
2  2 sp0000001-mRNA-1           
3  3 sp0000001-mRNA-1           
4  4 sp0000005-mRNA-1 GO:0003723
5  5 sp0000006-mRNA-1 GO:0016021
6  6 sp0000006-mRNA-1 GO:0016021
7  7 sp0000006-mRNA-2 GO:0016021
8  9 sp0000006-mRNA-3 GO:0016021

Answer 2

尝试使用数据框或数据表：

interproscan <- data.frame(interproscan)

unique(interproscan)

输出：

                V1        V14
1 sp0000001-mRNA-1           
4 sp0000005-mRNA-1 GO:0003723
5 sp0000006-mRNA-1 GO:0016021

示例数据：

require(data.table)
interproscan <- fread("V1,               V14
                       sp0000001-mRNA-1,           
                       sp0000001-mRNA-1,          
                       sp0000001-mRNA-1,            
                       sp0000005-mRNA-1, GO:0003723
                       sp0000006-mRNA-1, GO:0016021
                       sp0000006-mRNA-1, GO:0016021")

如何使用具有空白/缺失值的唯一功能

2 个答案: