Question

我有一个包含5列的数据框，其中4列是“正常”值，第5列是一组用定界符分隔的键/值对，格式为

key1 = value1 | key2 = value2 | key3 = value3

单行中没有重复的键，但是不能保证每行中键的顺序相同或键/值对的数量在每一行中都相同。

最简单的方法是将值分成约15列（以其键正确命名）？

Answer 1

这是使用tidyr，dplyr和stringr保持非键值列完整的一种选择。

创建测试数据

testdf <- data_frame(col1 = paste0("a", 1:4), col2 = paste0("b", 1:4), 
              col3 = paste0("c", 5:8),
              col4 = paste0("enst", 12099:12102),
              col5 = c("bang=1 | dong=2 | foo=6", "bang=2 | dong=3 | zups=foo", 
                       "dimbo=2 | bang=6 | jumbo=7 | zoo=camel", "boo=0"))

代码

library(stringr)
library(tidyr)
library(dplyr)

KVsep <- fixed(" | ")  #key-value separator
Vsep <- fixed("=")     #value separator

newDF <-  testdf %>%
  mutate(KVpairs = str_split(col5, KVsep)) %>%
  unnest(KVpairs) %>%
  separate(KVpairs, into = c("key", "value"), Vsep) %>%
  spread(key, value) %>%
  select(-col5)

结果

> newDF
# A tibble: 4 x 12
  col1  col2  col3  col4      bang  boo   dimbo dong  foo   jumbo zoo   zups 
* <chr> <chr> <chr> <chr>     <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr>
1 a1    b1    c5    enst12099 1     NA    NA    2     6     NA    NA    NA   
2 a2    b2    c6    enst12100 2     NA    NA    3     NA    NA    NA    foo  
3 a3    b3    c7    enst12101 6     NA    2     NA    NA    7     camel NA   
4 a4    b4    c8    enst12102 NA    0     NA    NA    NA    NA    NA    NA
>

Answer 2

好吧，我不认为有任何专门的功能，但是进行一些编程并不复杂。

c2nv <- function(x){
  # part 1 - convert vector to name-value pairs
  x <- strsplit(x, "|", fixed=TRUE) # split at |, then split at =
  x <- sapply(x,  function(.) t(trimws(sapply(lapply(., strsplit, "="), "[[", 1))))
  x1 <- lapply(x, function(.) setNames( .[,2], .[,1]))
  # part 2 - find all existing names in all rows
  u <- sort(unique(unlist(lapply(x1, names))))  # unique keys
  res <- t(sapply(x1, function(.) .[u]))   # look at each key in each row
  colnames(res) <- u
  res
}

样本数据

col <- c("bang=1 | dong = 2 | foo = 6", "bang1=2 | dong = 3 | zups = foo", 
"dimbo = 2 | bang = 6 | jumbo = 7 | zoo = camel", "boo = 0")

处理样本数据

> c2nv (col)
     bang bang1 boo dimbo dong foo jumbo zoo     zups 
[1,] "1"  NA    NA  NA    "2"  "6" NA    NA      NA   
[2,] NA   "2"   NA  NA    "3"  NA  NA    NA      "foo"
[3,] "6"  NA    NA  "2"   NA   NA  "7"   "camel" NA   
[4,] NA   NA    "0" NA    NA   NA  NA    NA      NA   
>

将分离的键/值对的dataframe列拆分为新列

2 个答案:

样本数据

处理样本数据