为什么不使用 ifelse() 基于两个 df 中的匹配观察值创建新列适用于此数据？

Question

为了先发制人，我已经使用

left_join()

（或

merge()

）找到了解决我的问题的方法，但我不太明白为什么

ifelse()

不适用于这个问题。我很想听听我可以做到这一点或改进我的

left_join()

使用的任何其他方法。抱歉，如果这是一篇很长的文章。

基本上，我试图通过将列

df1

中的观察结果与索引 df

df1$code

中相应列中的观察结果进行匹配，在数据框

index.df$code

中创建新列。新列

df1$type

将是

index.df$type

中对应于

df1$code

值的值：

#index data frame
index.df <- data.frame(
  code = c("c10", "c20", "c03", "c48", "c19"),
  id = c("apple", "strawberry", "pear", "banana", "blackberry"),
  type = c("pome", "aggregate", "pome", "berry", "aggregate")
)
> index.df
  code         id      type
1  c10      apple      pome
2  c20 strawberry aggregate
3  c03       pear      pome
4  c48     banana     berry
5  c19  blackberry aggregate

#df to add col to
df1 <- data.frame(
  code = c("c10", "c19", "c03", "c20", "c19", "c10", "c48", "c03", "c10", "c03"),
  id = c("apple", "blackberry", "pear","strawberry", "blackberry", "apple", "banana", "pear", "apple", "pear")
)
> df1
   code         id
1   c10      apple
2   c19 blackberry
3   c03       pear
4   c20 strawberry
5   c19 blackberry
6   c10      apple
7   c48     banana
8   c03       pear
9   c10      apple
10  c03       pear

这就是所需的输出

> df2
   code         id      type
1   c10      apple      pome
2   c19 blackberry aggregate
3   c03       pear      pome
4   c20 strawberry aggregate
5   c19 blackberry aggregate
6   c10      apple      pome
7   c48     banana     berry
8   c03       pear      pome
9   c10      apple      pome
10  c03       pear      pome

我尝试过

ifelse()

这样：

df2 <- df1 %>%
  mutate(df1, type = ifelse(df1$code == index.df$code, index.df$type, NA))

> df2
   no code         id      type
1   1  c10      apple      pome
2   2  c19 blackberry      <NA>
3   3  c03       pear      pome
4   4  c20 strawberry      <NA>
5   5  c19 blackberry aggregate
6   6  c10      apple      pome
7   7  c48     banana      <NA>
8   8  c03       pear      pome
9   9  c10      apple      <NA>
10 10  c03       pear      <NA>

为什么是这样的输出？我是否错误地使用了

ifelse()

？预先感谢您！

此外，我用来获得所需输出的（相当庞大的）代码是：

df1 <- data.frame(
  no = 1:10,
  code = c("c10", "c19", "c03", "c20", "c19", "c10", "c48", "c03", "c10", "c03"),
  id = c("apple", "blackberry", "pear","strawberry", "blackberry", "apple", "banana", "pear", "apple", "pear")
)

df2 <- index.df %>%
  left_join(df1, by = c("code", "id")) %>%
  arrange(no) %>%
  select(-no)

Answer 1

你在做

c("c10", "c19", "c03", "c20", "c19", "c10", "c48", "c03", "c10", "c03") == c("c10", "c20", "c03", "c48", "c19") 

看起来它是有效的，但可能不会像你想象的那样。

这也许更简洁一点？

df2 <- df1
df2$type <- index.df$type[match(df1$code, index.df$code)]