在r中的单词和分号之间拆分字符串答案

【问题标题】：Split string bewteen words and semicolon in r在r中的单词和分号之间拆分字符串
【发布时间】：2020-02-14 20:03:00
【问题描述】：

我在名为text 的列中有一个包含字符串的数据框（或数据表），如下所示：

text
name: john; surname: smith; age: 35; gender:male
name: mark; age:50
name: jack; surname: brown
name: tom; surname: travis; gender: male

如何将字符串的每个部分提取到同一数据框中的单独列中？我想要以下列：

name.text
name: john
name: mark
name: jack
name: tom

surname.text
surname: smith
(empty)
surname: brown
surname: travis

age.text
age: 35
age: 50
(empty)
(empty)

gender.text
gender:male
(empty)
(empty)
gender:male

谢谢！

【问题讨论】：

标签： r string dataframe parsing data.table

【解决方案1】：

这是一种选择，我们将; 处的元素拆分，然后将separate 分成两列，并从“长”格式重新调整为“宽”格式

library(dplyr)
library(tidyr)
library(stringr)
library(tibble)
df1 %>%
   rownames_to_column('rn') %>%
   separate_rows(text, sep = ';\\s*') %>%
   separate(text, into = c('key', 'val'), sep=":\\s*") %>% 
   pivot_wider(names_from = key, values_from = val, 
            values_fill = list(val = "(empty)")) %>%
   select(-rn) %>%
   imap_dfr(~ case_when(.x != "(empty)" ~ str_c(.y, .x, sep=":"), TRUE ~ .x)) %>%
   rename_all(~ str_c(., ".text"))
# A tibble: 4 x 4
#  name.text surname.text   age.text gender.text
#  <chr>     <chr>          <chr>    <chr>      
#1 name:john surname:smith  age:35   gender:male
#2 name:mark (empty)        age:50   (empty)    
#3 name:jack surname:brown  (empty)  (empty)    
#4 name:tom  surname:travis (empty)  gender:male

或者使用base R

nm1 <- c("name", "surname", "age", "gender")
lst1 <-  lapply(strsplit(df1$text, ";\\s*"), function(x) {
          prfx <- sub(":.*", "", x)
          x1 <- x[match(nm1, prfx)]
          replace(x1, is.na(x1), "(empty)")})
out <- do.call(rbind.data.frame, lst1)
names(out) <- paste0(nm1, ".text")
out
#   name.text    surname.text age.text  gender.text
#1 name: john  surname: smith  age: 35  gender:male
#2 name: mark         (empty)   age:50      (empty)
#3 name: jack  surname: brown  (empty)      (empty)
#4  name: tom surname: travis  (empty) gender: male

数据

df1 <- structure(list(text = c("name: john; surname: smith; age: 35; gender:male", 
"name: mark; age:50", "name: jack; surname: brown", 
"name: tom; surname: travis; gender: male"
)), class = "data.frame", row.names = c(NA, -4L))

【讨论】：

非常感谢您的回复。我需要使用几乎基本的 R （因为其他项目依赖，不要问为什么:)）。你能帮忙吗，如何仅使用基础 R 来解决这个问题？
@Makaroni。在帖子中，你指定了data.table（所以你在使用那个包）
@Makaroni。更新为base R

【解决方案2】：

d = lapply(strsplit(df1$text, "; ?"), function(x){
    data.frame(do.call(rbind, strsplit(x, ": ?")), stringsAsFactors = FALSE)
})

fields = unique(unlist(lapply(d, function(x) x$X1)))

d2 = do.call(rbind, lapply(d, function(x)
    data.frame(fields, val = x$X2[match(fields, x$X1)])))

d2[order(match(d2$fields, fields)),]

【讨论】：