无法使用 gsub 或 trimws 删除前导/尾随空格

Question

我正在尝试处理需要大量清理的数据集。我有一个主题名称，我似乎无法从中删除前导空格。

示例数据：

Data <- dput(Data)
structure(list(Teacher = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L
), .Label = c("Please.rate.teacher:.JOHN.DOE .Overall.rating.for.teacher", 
"Please.rate.teacher: Jane.Doe.Overall.rating.for.teacher"), class = "factor"), 
    Overall_Rating = c(5L, 4L, 5L, 4L, 4L, 5L, 4L, 4L, 4L, 4L, 
    3L, 5L, 4L, 4L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 3L)), .Names = c("Teacher", 
"Overall_Rating"), class = "data.frame", row.names = c(NA, -22L
))

我的清洁尝试：

Data_clean <- Data %>%
 mutate(Teacher = as.character(Teacher),
Teacher = gsub("Please.rate.teacher|.Overall.rating.for.teacher|[:]", "", Teacher),
            Teacher = gsub("[.]", " ", Teacher),
            Teacher = trimws(Teacher),
            Teacher = tolower(Teacher), Teacher = tools::toTitleCase(Teacher)) 

导致剩余的前导和尾随空格，这也打破了第二个名称的标题大小写：

unique(Data_clean$Teacher)
[1] "John Doe " " jane Doe"

第一个名字仍然有尾随空格，第二个名字有前导空格。

我怎样才能删除它？

Answer 1

我怀疑您的数据包含非 ASCII 空格，例如 "\u00A0"。 trimws 函数只会删除 ASCII 空格字符。

尝试运行utf8::utf8_print(unique(Data_clean$Teacher), utf8 = FALSE) 看看是否是这种情况。

要处理非 ASCII 空格，请将代码中的 trimws(x) 替换为

gsub("(^[[:space:]]*)|([[:space:]]*$)", "", x)

Answer 2

这是一个完全可重复的示例，特别是 stringr 和 str_trim，因为我不知道为什么 trimws 不适合您。您发布的代码给了我相同的输出，正确地将大小写更改为标题并删除了空格。

data <- structure(list(Teacher = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 
                                     1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L
), .Label = c("Please.rate.teacher:.JOHN.DOE .Overall.rating.for.teacher", 
              "Please.rate.teacher: Jane.Doe.Overall.rating.for.teacher"), class = "factor"), 
Overall_Rating = c(5L, 4L, 5L, 4L, 4L, 5L, 4L, 4L, 4L, 4L, 
                   3L, 5L, 4L, 4L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 3L)), .Names = c("Teacher", 
                                                                                "Overall_Rating"), class = "data.frame", row.names = c(NA, -22L
                                                                                ))

library(tidyverse)
data %>%
  mutate(
    Teacher = Teacher %>%
      str_remove_all("Please.rate.teacher:|.Overall.rating.for.teacher") %>%
      str_replace_all("\\.", " ") %>%
      str_trim() %>%
      str_to_title()
  ) %>%
  `[[`(1) %>%
  unique()
#> [1] "John Doe" "Jane Doe"

由reprex package (v0.2.0) 于 2018 年 3 月 15 日创建。

Answer 3

这个怎么样？

Data_clean <- Data %>%
     mutate(Teacher = gsub("Please.rate.teacher|\\s*\\.Overall.rating.for.teacher|:", "", Teacher),
            Teacher = gsub("\\.", " ", Teacher),
            Teacher = trimws(Teacher),
            Teacher = tolower(Teacher), Teacher = tools::toTitleCase(Teacher))

unique(Data_clean$Teacher);
#[1] "John Doe" "Jane Doe"

说明：替换 Teacher 中 ".Overall.rating..." 之前出现的可选 (>=0) 空格。

---

样本数据

Data <- structure(list(Teacher = structure(c(1L, 1L, 1L, 1L, 1L, 1L,
1L, 1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L, 2L
), .Label = c("Please.rate.teacher:.JOHN.DOE .Overall.rating.for.teacher",
"Please.rate.teacher: Jane.Doe.Overall.rating.for.teacher"), class = "factor"),
    Overall_Rating = c(5L, 4L, 5L, 4L, 4L, 5L, 4L, 4L, 4L, 4L,
    3L, 5L, 4L, 4L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 3L)), .Names = c("Teacher",
"Overall_Rating"), class = "data.frame", row.names = c(NA, -22L
))