用R中的regexp解析(可能)非existend数字

时间:2015-12-13 12:05:11

标签: regex r

我正在尝试使用包stringr从R中的字符串中提取数字。有时,数字不存在。以下是一些示例字符串:

str <- c(
"cash dividends per share $ - $ - $ - $ 0.08 $ 0.16 cash",
"cash dividends per share $ 0.01 $ 12.10 $ 0.01 $ 0.08 $ 0.16 hello",
"cash dividends per share $ - $ - $ 0.91 $ - $ 0.16 world",
"cash dividends per share - - 0.12 - 0.16 hsac",
"cash dividends per share $ - $ - $ - $ - $ 0.16 afterwards",
"cash dividends per share $0.12 $ - $0.1 $ - $ - comes",
"cash dividends per share 0.12 - 0.12 - - text",
"cash dividends per share... 0.12 - 0.12 - - random",
"cash dividends per share...0.123 0.321 - - 0.12 blu",
"cash dividends per share ..... $ 0.12 $ - $ 0.12  $ - $ - foo",
"cash dividends per share ..... $0.42 $0.42 $-  $- $- bar")

我构建了以下正则表达式,IMO应该匹配所有情况,但它没有。当然,我也尝试了各种不同的,但我无法找到正确的一个(我甚至无法看到我提出的问题):

library("stringr")
rgxp <- "cash dividends [declared]* per share[ \\.]+[\\$]?[ ]?([-0.9\\.]+)[ ]?[\\$]?[ ]?([-0.9\\.]+)[ ]?[\\$]?[ ]?([-0.9\\.]+)[ ]?[\\$]?[ ]?([-0.9\\.]+)[ ]?[\\$]?[ ]?([-0.9\\.]+).*"
str_match_all(str, rgxp)

您是否看到上述正则表达式失败的原因?

编辑:我应该说我想要的输出是 alyways 一个包含五个元素的向量,即如果没有数字,则为数字或连字符。谢谢!

2 个答案:

答案 0 :(得分:3)

使用gsub用空格替换任意两个或多个连续点以及任何不是减号,数字或点的字符,然后使用read.table读入。如果希望减号出现在NA所在的位置,请忽略na.strings = "-"。没有包使用。

DF <- read.table(text = gsub("[^-0-9.]+|\\.{2,}", " ", str), fill = TRUE, na.strings = "-")

给出这个data.frame:

> DF
      V1     V2   V3   V4   V5
1     NA     NA   NA 0.08 0.16
2  0.010 12.100 0.01 0.08 0.16
3     NA     NA 0.91   NA 0.16
4     NA     NA 0.12   NA 0.16
5     NA     NA   NA   NA 0.16
6  0.120     NA 0.10   NA   NA
7  0.120     NA 0.12   NA   NA
8  0.120     NA 0.12   NA   NA
9  0.123  0.321   NA   NA 0.12
10 0.120     NA 0.12   NA   NA
11 0.420  0.420   NA   NA   NA

注意:如果您想用零替换NA,请使用:DF[is.na(DF)] <- 0

答案 1 :(得分:2)

如果你想捕捉数字和连字符:

rgxp <- "([0-9]+\\.?[0-9]*)|(-)"
str_extract_all(str, rgxp)

[[1]]
[1] "-"    "-"    "-"    "0.08" "0.16"

[[2]]
[1] "0.01"  "12.10" "0.01"  "0.08"  "0.16" 

[[3]]
[1] "-"    "-"    "0.91" "-"    "0.16"

[[4]]
[1] "-"    "-"    "0.12" "-"    "0.16"

[[5]]
[1] "-"    "-"    "-"    "-"    "0.16"

[[6]]
[1] "0.12" "-"    "0.1"  "-"    "-"   

[[7]]
[1] "0.12" "-"    "0.12" "-"    "-"   

[[8]]
[1] "0.12" "-"    "0.12" "-"    "-"   

[[9]]
[1] "0.123" "0.321" "-"     "-"     "0.12" 

[[10]]
[1] "0.12" "-"    "0.12" "-"    "-"   

[[11]]
[1] "0.42" "0.42" "-"    "-"    "-"   

如果您只想要数字:

rgxp <- "[0-9]+\\.?[0-9]*"
str_extract_all(str, rgxp)

[[1]]
[1] "0.08" "0.16"

[[2]]
[1] "0.01"  "12.10" "0.01"  "0.08"  "0.16" 

[[3]]
[1] "0.91" "0.16"

[[4]]
[1] "0.12" "0.16"

[[5]]
[1] "0.16"

[[6]]
[1] "0.12" "0.1" 

[[7]]
[1] "0.12" "0.12"

[[8]]
[1] "0.12" "0.12"

[[9]]
[1] "0.123" "0.321" "0.12" 

[[10]]
[1] "0.12" "0.12"

[[11]]
[1] "0.42" "0.42"