创建一个包含grepped字符串的新列

Posted 2021-04-03

我有3500多个项目的数据框，我想在Project_Description专栏中搜索40个关键词。如果Project_Description包含一个或多个关键字，我想创建一个新列，并使用关键字标记项目的行。

如何创建循环遍历关键字的if语句，如果找到关键字，则用关键字标记正确的行？特别是如果Project_Description可能包含多个关键字？

到目前为止，我已经能够提取出包含Project_Description列中至少一个关键词的项目行​​。

key_words <- c("who","what","when","where","why", etc...)

dataframe_key_words <- c()

for (i in 1:length(key_words)){
dataframe_key_words <- rbind(dataframe_key_words, dataframe_original[grep(key_words[i], dataframe_original$Project_Description), ]
}

答案

你可以试试这个：

library(data.table)
library(stringi)
key_words <- c("where", "why")
pat <- paste0("(", paste0(key_words, collapse = "|"), ")")
DT <- data.table(descr = c("where is the sample data? why do you do this?", 
                           "this doesn't have any of the keywords"))
DT[, kw := lapply(stri_match_all_regex(descr, pat), function(x) x[, 2])][]

#                                            descr        kw
# 1: where is the sample data? why do you do this? where,why
# 2:         this doesn't have any of the keywords        NA