3

我有一个大数据框，里面充满了以下字符：

 x <- c("Y188","Y204" ,"Y221","EP121_1" ,"Y233" , "Y248" ,"Y268", "BB2","BB20",  
 "BB32" ,"BB044" ,"BB056" , "Y234" , "Y249" ,"Y271" ,"BB3", "BB21", "BB33",
 "BB045","BB057" ,"Y236", "Y250", "Y272" , "BB4", "BB22" )

如您所见，某些标签（例如 BB20）只有两个整数。我希望整个字符列表至少有 3 个这样的整数（如果有帮助，问题只在 BB 标签中）：

Y188, Y204, Y221, EP121_1, Y233, Y248, Y268, BB002, BB020, BB032, BB044,
BB056, Y234, Y249, Y271, BB003, BB021, BB033, BB045, BB057, Y236, Y250,
Y272, BB004, BB022

我研究了sprintfandFormatC功能，但仍然没有运气。

4

4 回答 4

3

gsub嵌套调用的有力方法：

gsub("(.*[A-Z])(\\d{1}$)", "\\100\\2",
     gsub("(.*[A-Z])(\\d{2}$)", "\\10\\2", x))
# [1] "Y188"    "Y204"    "Y221"    "EP121_1" "Y233"    "Y248"    "Y268"    "BB002"   "BB020"  
# [10] "BB032"   "BB044"   "BB056"   "Y234"    "Y249"    "Y271"    "BB003"   "BB021"   "BB033"  
# [19] "BB045"   "BB057"   "Y236"    "Y250"    "Y272"    "BB004"   "BB022"
于 2016-02-12T23:03:46.577 回答
2

肯定有一种更通用的方法可以做到这一点，但对于这样一个本地化的任务，两个简单的sub就足够了：为两位数添加一个尾随零，为一位数添加两个尾随零。

x <- sub("^BB(\\d{1})$","BB00\\1",x)
x <- sub("^BB(\\d{2})$","BB0\\1",x)
于 2016-02-12T23:05:08.337 回答
1

这可行，但会有边缘情况

# indicator for numeric of length less than three
num <- gsub("[^0-9]", "", x)
id <- nchar(num) < 3

# overwrite relevant values with the reformatted ones
x[id] <- paste0(gsub("[0-9]", "", x)[id],
                     formatC(as.numeric(num[id]), width = 3, flag = "0"))

 [1] "Y188"    "Y204"    "Y221"    "EP121_1" "Y233"    "Y248"    "Y268"    "BB002"   "BB020"   "BB032"  
[11] "BB044"   "BB056"   "Y234"    "Y249"    "Y271"    "BB003"   "BB021"   "BB033"   "BB045"   "BB057"  
[21] "Y236"    "Y250"    "Y272"    "BB004"   "BB022"
于 2016-02-12T23:20:30.897 回答
0

可以使用sprintf和gsub函数来完成。这一步将提取数值并更改其格式。

num=sprintf("%03d",as.numeric(gsub("[^[:digit:]]", "", x)))

下一步是粘贴更改格式的数字

x=paste(gsub("[^[:alpha:]]", "", x),num,sep="")

于 2016-02-13T12:55:04.190 回答