Question

假设A和B是数据集中的列，我想开发一种模糊匹配逻辑-如果A列中至少有一个单词与B列中的一个单词匹配，除了单词“ bank”和“ of”，我们在新列中分配1，如果匹配项为0，则我们分配0。我想在R中做到这一点。

A                          B
BANK OF AMERICA         CHASE BANK
BANK OF AMERICA         BANK OF AMERICA, N.A.
BANK OF HOPE            HOPE BANK
T.D BANK                CHASE BANK

预期产量

A                         B                        C
BANK OF AMERICA         CHASE BANK                 0
BANK OF AMERICA         BANK OF AMERICA, N.A       1
BANK OF HOPE            HOPE BANK                  1
T.D. BANK               CHASE BANK                 0

Answer 1

这是另一个选项-带有dplyr和stringr。

df <- data.frame(A = c(rep("BANK OF AMERICA", 2), "BANK OF HOPE", "T.D BANK"),
                 B = c("CHASE BANK", "BANK OF AMERICA, N.A.", "HOPE BANK", "CHASE BANK"),
                 stringsAsFactors = FALSE) 


df <- df %>% 
  mutate(C = str_remove_all(B, c("BANK|OF|,")), #remove stopwords
         C = str_trim(C), #remove whitespace from start/end
         C = str_replace_all(C, "  ", ""), #remove double whitespaces
         C = str_replace_all(C, " ", "|")) %>% #replace whitespace with |
  mutate(D = as.numeric(str_detect(A, C))) %>% 
  select(A, B, D)

                A                     B D
1 BANK OF AMERICA            CHASE BANK 0
2 BANK OF AMERICA BANK OF AMERICA, N.A. 1
3    BANK OF HOPE             HOPE BANK 1
4        T.D BANK            CHASE BANK 0

Answer 2

也许以下基本R选项可能有帮助

df$C <- +do.call(
  function(...) mapply(function(...) any(!intersect(...) %in% c("BANK","OF")),...),
  Map(function(x) strsplit(x,"[[:punct:][:blank:]]",perl = TRUE), df, USE.NAMES = FALSE)
)

给出

> df
                A                     B C
1 BANK OF AMERICA            CHASE BANK 0
2 BANK OF AMERICA BANK OF AMERICA, N.A. 1
3    BANK OF HOPE             HOPE BANK 1
4        T.D BANK            CHASE BANK 0

Answer 3

我相信正则表达式和apply的结合在这里很有效。

> df <- data.frame(A = c('BANK OF AMERICA', 'BANK OF AMERICA', 'BANK OF HOPE', 'T.D BANK'),
                 B = c('CHASE BANK', 'BANK OF AMERICA, N.A.', 'HOPE BANK', 'CHASE BANK'),
                 stringsAsFactors = FALSE)

> f <- function(x) {
  left <- strsplit(x[1], "(BANK OF\\s|\\s|,|\\sBANK)")[[1]]
  right <- strsplit(x[2],  "(BANK OF\\s|\\s|,|\\sBANK)")[[1]]
  ans <- left %in% right
  as.integer(all(ans[!(left %in% "")]))
}

> df$C <- apply(df, 1, f)
> df
                A                     B C
1 BANK OF AMERICA            CHASE BANK 0
2 BANK OF AMERICA BANK OF AMERICA, N.A. 1
3    BANK OF HOPE             HOPE BANK 1
4        T.D BANK            CHASE BANK 0

如果一列包含另一列的单词匹配

3 个答案: