R：使用 gsub 和 grep 从文本列表创建新数据框

Question

如何从文本列表中提取重要信息 - 姓名、年龄、（疾病）和体重并创建新的数据列表或框架？

test<-c("James is approximately age 25 & 26 weighted 130lbs", 
        "Angelina is age 40 (Diabetes)", 
        "Harry Peterson is male with ages 27")

这样做

我能够在括号内对名称/和疾病进行子集化。

> sapply(strsplit(test, "\s+"),"[",1)
[1] "James"    "Angelina" "Harry"

> gsub("[\(\)]","\1", regmatches(test, gregexpr("(?<=\().*?(?=\))",test, perl=T)))
[1] "character0" "Diabetes"   "character0"

然而， 无法对年龄 25 和 26 进行子集化，并且 grep 'ages'

> paste(grep(pattern="age",  trimws(strsplit(test, " ")[[1]]), value = TRUE),as.numeric(sub(".*age. 
(\d+).*", "\1", test[[1]])) )
[1] "age 25"

如何从文本中提取所有数字和符号？ 像 "age 25 & 26"

如何设置提取年龄和年龄的年龄模式？ 年龄 27 -> "age 27"，加权 -> "weight 130"

我怎样才能按照下面的顺序 grep 所有信息，而不是单独的子集名称、年龄、体重和 ()？

 c("James","age 25 & 26", "weight 130", "Angelina","age 40", "Diabetes", "Harry", "age 27")

并最终创建如下数据框

          age        weight  illness   
James     "25 & 26"  "130"   NA        
Angelina  "40"       NA     "Diabetes"
Harry     "27"       NA      NA

如果您只能部分回答，这也会有所帮助。谢谢。

Answer 1

gsub(" .*", "", test)
# [1] "James"    "Angelina" "Harry"   

trimws(gsub("ages?", "", regmatches(test, gregexpr("ages?\s*[-&0-9 ]+\b", test, perl = TRUE))))
# [1] "25 & 26" "40"      "27"     

weights <- regmatches(test, gregexpr("weight(s|ed)? [0-9]+(lb|pound|kg|g)?", test))
weights[lengths(weights) < 1] <- NA_character_
trimws(gsub("weight(s|ed)?", "", unlist(weights)))
# [1] "130lb" NA      NA     

ill <- regmatches(test, gregexpr("(?<=\().*(?=\))", test, perl = TRUE))
ill[lengths(ill) < 1] <- NA_character_
unlist(ill)
# [1] NA         "Diabetes" NA

R：使用 gsub 和 grep 从文本列表创建新数据框

R: Creating new data frame from text list using gsub, and grep

text

r

gsub

dataframe