将字符串拆分为多个固定宽度的列答案

【问题标题】：Splitting strings into multiple fixed width columns将字符串拆分为多个固定宽度的列
【发布时间】：2019-03-02 16:22:44
【问题描述】：

我正在尝试使用 str_split 将以下观察结果拆分为特定格式。

"00010943900008" "00010946803119" "00010946803219" "00010946803219" "00010946803219" "00010948700007"

我正在尝试将其拆分为不同的列。

所以第一个观察结果如下所示：

Column x = 00

Column y = 01

Column z = 09439

Column w = 00008

其中 x 列将始终是观察中的前 2 个数字，y 列将是接下来的 2 个数字，z 列将是接下来的 5 个数字，w 列将是最后 5 个数字

数据

string <- c("00010943900008", "00010946803119", "00010946803219", "00010946803219", 
"00010946803219", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00010948700007", "00010948700007", 
"00010948700007", "00010948700007", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016", 
"00011820000016", "00011820000016", "00011820000016", "00011820000016"
)

【问题讨论】：

我建议将其写入文件，然后使用read.fwf 将其读回。否则可能使用substr。

标签： r regex string split

【解决方案1】：

您可以使用\n 作为分隔符连接您的数据或将其写入文件，然后使用readr::read_fwf 或read.fwf（仅来自文件）将其作为固定宽度格式导入。这是没有写入磁盘的readr::read_fwf 版本：

library(readr)
result = read_fwf(paste(string, collapse = "\n"),
                  col_positions = fwf_widths(c(2, 2, 5, 5), col_names = c("x", "y", "z", "w")))
head(result)
# # A tibble: 6 x 4
#   x     y     z     w
#   <chr> <chr> <chr> <chr>
# 1 00    01    09439 00008
# 2 00    01    09468 03119
# 3 00    01    09468 03219
# 4 00    01    09468 03219
# 5 00    01    09468 03219
# 6 00    01    09487 00007

【讨论】：

将textConnection 与read.fwf 一起使用将使用字符串：read.fwf(textConnection(string), widths = c(2,2,5,5), col_names = c("x", "y", "z", "w")。

【解决方案2】：

来自tidyr 的extract。 extract 将每个正则表达式捕获组转换为自己的列。如果不想保留原来的列，可以设置remove = TRUE（默认）：

library(dplyr)
library(tidyr)

string %>%
  data.frame(string = .) %>%
  extract(string, c("x","y","z","w"), "^(\\d{2})(\\d{2})(\\d{5})(\\d{5})", remove = FALSE)

输出：

            string  x  y     z     w
1   00010943900008 00 01 09439 00008
2   00010946803119 00 01 09468 03119
3   00010946803219 00 01 09468 03219
4   00010946803219 00 01 09468 03219
5   00010946803219 00 01 09468 03219
6   00010948700007 00 01 09487 00007
7   00010948700007 00 01 09487 00007
8   00010948700007 00 01 09487 00007
9   00010948700007 00 01 09487 00007
10  00010948700007 00 01 09487 00007
11  00010948700007 00 01 09487 00007
12  00010948700007 00 01 09487 00007

【讨论】：

【解决方案3】：

您可以从字符串创建一个数据框，然后使用 substr()，它根据位置返回部分字符串：

data<- as.data.frame(string)
data$x <- substr(string,1,2)
data$y <- substr(string,3,4)
data$z <- substr(string,5,9)
data$w <- substr(string,10,14)

【讨论】：

【解决方案4】：

我们可以使用regex和read.table（这只有在模式相同的情况下才有效）：

> read.table(text=gsub("(\\d{2})(\\d{2})(\\d{5})(\\d{5})", "\\1,\\2,\\3,\\4", string),
             colClasses="character", sep=",", stringsAsFactors = FALSE)
    V1 V2    V3    V4
1   00 01 09439 00008
2   00 01 09468 03119
3   00 01 09468 03219
4   00 01 09468 03219
5   00 01 09468 03219
6   00 01 09487 00007
7   00 01 09487 00007
8   00 01 09487 00007
9   00 01 09487 00007
10  00 01 09487 00007
...

【讨论】：

【解决方案5】：

使用 tidyr::separate:

library(tidyr)

data.frame(string = string[1:5]) %>% 
  separate(string, c("x", "y", "z", "w"),
           sep = c(2, 4, 9), remove = FALSE)

#           string  x  y     z     w
# 1 00010943900008 00 01 09439 00008
# 2 00010946803119 00 01 09468 03119
# 3 00010946803219 00 01 09468 03219
# 4 00010946803219 00 01 09468 03219
# 5 00010946803219 00 01 09468 03219

【讨论】：