如何检查数据框的列的每个单元格（列表）在R中是否唯一？

academic.user 发表于 Dev

学术用户

我有一个非常大的数据，我想做的是检查列表是否具有一个字符串多个值：请考虑以下数据帧：

df1 <- structure(list(id = 1:3, 
                      book_id = c("[\"19167120\",\"book\", \"237494310\",\"195166798\",\"book\",\"book.a\"]", 
                                  "[\"19167120\",\"237494310\",\"story\",\"book\",\"19167120\"]", "[]")), 
                 .Names = c("id", "book_id"), 
                 class = "data.frame", 
                 row.names = c(NA, -3L))

这是：

     id                                                      book_id
1  1 ["19167120","book", "237494310","195166798","book","book.a"]
2  2           ["19167120","237494310","story","book","19167120"]
3  3                                                           []

我想做的是检查book_id是否有任何列表值具有多个字符串值，例如，在这里：字符串值"19167120"对于第二行重复。并为每个单元格提取它们，然后从每个单元格中将其删除

输出：两个单独的数据框：

   id                                                          book_id      duplicate
1:  1 ["19167120", "book", "237494310", "195166798", "book", "book.a"]     "book"
2:  2           ["19167120", "237494310", "story", "book", "19167120"] "19167120"
3:  3                                                               []        0

  id                                                  book_id
1:  1 ["19167120", "book", "237494310", "195166798", "book.a"]
2:  2               ["19167120", "237494310", "story", "book"]
3:  3                                                       []

我知道我应该使用anyduplicated()和unique()为了得到我的答案，但是我在他们周围工作，无法解决问题。

编辑：Gregor的第一个建议将是这样，但是如果有人能像我先解释的那样帮助我获得输出，我将不胜感激，

id     book_id
1:  1  "19167120"
2:  1 "237494310"
3:  1 "195166798"
4:  2  "19167120"
5:  2 "237494310"
6:  2  "19167120"
> unique(df1)
   id     book_id
1:  1  "19167120"
2:  1 "237494310"
3:  1 "195166798"
4:  2  "19167120"
5:  2 "237494310"
> duplicated(df1)
[1] FALSE FALSE FALSE FALSE FALSE  TRUE

A5C1D2H2I1M1N2O1R2T1

这是一种替代方法，与从“长”数据集开始并从那里开始的想法有关。

这是您的长数据集。

library(splitstackshape)
x <- cSplit(df1, "book_id", ",", "long")[, book_id := gsub(
    "[][]", "", book_id)]

在这里，我们添加带有重复值的“重复”列：

x[, duped := paste(unique(book_id[duplicated(book_id)], 
                   collapse = ", ")), by = id]

现在，我们可以轻松创建您的第一个所需的输出：

dupedX <- x[, list(book_id = sprintf("[%s]", paste(book_id, collapse = ", ")),
                   duped = paste(unique(duped), collapse = ", ")), by = id]
dupedX
#    id                                book_id      duped
# 1:  1 ["19167120", "237494310", "195166798"]         NA
# 2:  2  ["19167120", "237494310", "19167120"] "19167120"
# 3:  3                                     []         NA

还有您的第二个：

uniqueX <- x[, list(book_id = sprintf(
  "[%s]", paste(unique(book_id), collapse = ", "))), by = id]
uniqueX
#    id                                book_id
# 1:  1 ["19167120", "237494310", "195166798"]
# 2:  2              ["19167120", "237494310"]
# 3:  3                                     []

本文收集自互联网，转载请注明来源。

如有侵权，请联系[email protected] 删除。