How do I separate every character in a string in a vector into a column using tidyr

问题

I want to separate each string in the vector into columns but I can't do it!

library(tidyr)  
    library(dplyr)
    df <- data.frame(x = c("abe", "bas", "dds", "eer"))
    df %>% separate(x, c("A", "B", "C"), sep=1)

The output I want looks like this

  A B C
1 a b e
2 b a s
3 d d s
4 e e r

That sep=1 works for 2 characters but doesn't work for 3. I was hoping a regex like sep="." or sep="[a-z]" would work too but it doesn't.

This is probably super easy but I'm new to R. Won't someone please help!

回答1:

You were quite close with your own solution. Simply add a second position for the sep argument.

So:

library(tidyr)  
library(dplyr)
df <- data.frame(x = c("abe", "bas", "dds", "eer"))
df %>% separate(x, c("A", "B", "C"), sep = c(1,2))

  A B C
1 a b e
2 b a s
3 d d s
4 e e r

回答2:

Method 1

Use a positive lookbehind with separate:

df %>%
    separate(x, c("A", "B", "C"), sep = "(?<=.)", extra = "drop")
#  A B C
#1 a b e
#2 b a s
#3 d d s
#4 e e r

Note that this will only work if every string x consists of exactly three characters.

Method 2

Use strsplit:

df %>%
    mutate(tmp = strsplit(as.character(x), "")) %>%
    unnest() %>%
    group_by(x) %>%
    mutate(n = 1:n()) %>%
    spread(n, tmp) %>%
    ungroup() %>%
    select(-x)
## A tibble: 4 x 3
#  `1`   `2`   `3`
#  <chr> <chr> <chr>
#1 a     b     e
#2 b     a     s
#3 d     d     s
#4 e     e     r

This will also allow for strings x of varying lengths, by padding columns with NAs if necessary.

回答3:

In spite you want a non R-base solution, here's an R base approach just for the record.

> x <- data.frame(do.call(rbind, strsplit(as.character(df$x), "")))
> names(x) <- LETTERS[1:3]
> x
  A B C
1 a b e
2 b a s
3 d d s
4 e e r

来源：https://stackoverflow.com/questions/49887440/how-do-i-separate-every-character-in-a-string-in-a-vector-into-a-column-using-ti

标签

tidyr