Question

我有两个数据帧A和B. A有完整的句子，B有我正在寻找的经常性短语。我想找到A中的所有行，其中字符串/部分字符串存在于数据帧B中。例如，

Dataframe A有：

    "Sally is great"
     "John is great"
  "Sally likes peas"
 "John likes onions"
  "Jane is in Paris"
"Archie is in Paris"

Dataframe B有：

"in Paris"
"is great"

输出将是：

    "Sally is great"
     "John is great"
  "Jane is in Paris"
"Archie is in Paris"

因为这些是在数据帧B中存在字符串/子字符串的行。

相当于WHERE x LIKE＆＃39;％substring％＆＃39;在SQL中但是对于一组子字符串

我在A中接近200万行，在B中接近300,000行。我考虑过将str_match与循环一起使用，但考虑到数据大小，它可能不是一个可行的解决方案

Answer 1

一种方法是浏览较小集的元素，并使用grep检查它是否存在于较大的集合中。

big = c("Sally is great",
        "John is great",
        "Sally likes peas",
        "John likes onions",
        "Jane is in Paris",
        "Archie is in Paris")
small = c("in Paris",
          "is great")

big[unlist(lapply(small, function(a) grep(a, big)))]
#[1] "Jane is in Paris"   "Archie is in Paris" "Sally is great"     "John is great"

Answer 2

我们可以使用stri_detect

中的stringi

library(stringi)
big[stri_detect(big, regex = paste(small, collapse="|"))]
#[1] "Sally is great"     "John is great"      "Jane is in Paris"  
#[4] "Archie is in Paris"

数据

big <- c("Sally is great",
    "John is great",
    "Sally likes peas",
    "John likes onions",
    "Jane is in Paris",
    "Archie is in Paris")
small <- c("in Paris",
      "is great")

从数据框A中查找字符串，其子字符串存在于B中

2 个答案:

数据