如何过滤和复制 R 数据框中的值？

Question

我有一个如下所示的数据框

Test <- c('Sodium','Heamo Index.','Lipae Index','ictoric index','Chloride','Blood pressure','Test Index')
value <- c(12,'No haemo',NA,'No ict',21,32,NA)
TextualResults <- c('low value',NA,'NO LIPA',NA,'HIGH','low','NO TEST')

df_test <- data.frame(Test,value,TextualResults)

输入如下图所示

我正在尝试这样的事情

library(tidyverse)

df_test %>%  
  filter(stringr::str_detect(type, 'index|Index|INDEX'))

但不确定如何根据NA检查复制值并以优雅的方式使搜索不区分大小写

我希望我的输出如下所示

如您所见，在Test 列下，每当我们找到包含index（不区分大小写）的值时，我们必须确保这些行的value 列不是NA。

默认情况下，value 列下的 index 行可能具有值，但也可能是 NA。

因此，当它是 NA 时，我们会选择 TextualResults 列下存在的任何值，并将其放回 value 列中以获取包含 index 术语的行。对于包含index 术语的行，基本上value 列永远不会是NA

Answer 1

可以试试：

library(dplyr)

df_test %>%
  mutate_all(as.character) %>% # or mutate_at(-1, as.character) if you'd like to keep first as factor
  mutate(
    value = case_when(
      grepl('index', Test, ignore.case = TRUE) ~ coalesce(value, TextualResults),
      TRUE ~ value
    )
  )

输出：

            Test    value TextualResults
1         Sodium       12      low value
2    Heamo Index No haemo           <NA>
3    Lipae Index  NO LIPA        NO LIPA
4  ictoric index   No ict           <NA>
5       Chloride       21           HIGH
6 Blood pressure       32            low
7     Test Index  NO TEST        NO TEST

这应该适用于所有索引组合，无论它是在点、其他字符、大写/小写等之前还是之后。

考虑一下：

Test <- c('Sodium','Heamo Index.','Lipae Index..','ictoric __index','Chloride','Blood pressure','Test $$Index')
value <- c(12,'No haemo',NA,NA,21,32,NA)
TextualResults <- c('low value',NA,'NO LIPA','No ict','HIGH','low','NO TEST')

df_test <- data.frame(Test,value,TextualResults)

             Test    value TextualResults
1          Sodium       12      low value
2    Heamo Index. No haemo           <NA>
3   Lipae Index..     <NA>        NO LIPA
4 ictoric __index     <NA>         No ict
5        Chloride       21           HIGH
6  Blood pressure       32            low
7    Test $$Index     <NA>        NO TEST

使用上述代码输出：

             Test    value TextualResults
1          Sodium       12      low value
2    Heamo Index. No haemo           <NA>
3   Lipae Index..  NO LIPA        NO LIPA
4 ictoric __index   No ict         No ict
5        Chloride       21           HIGH
6  Blood pressure       32            low
7    Test $$Index  NO TEST        NO TEST

Answer 2

我会使用if_else() 来实现这一点。为了避免因素，我首先将它变成一个小标题而不是数据框。

Test <- c('Sodium','Heamo Index','Lipae Index','ictoric index','Chloride','Blood pressure','Test Index')
value <- c(12,'No haemo',NA,'No ict',21,32,NA)
TextualResults <- c('low value',NA,'NO LIPA',NA,'HIGH','low','NO TEST')
df_test <- tibble(Test,value,TextualResults)

contains_index <- str_detect(df_test$Test, 'index|Index|INDEX')

new_df <-
  df_test %>%
  mutate(value = if_else(is.na(value) & contains_index, TextualResults, value))

Answer 3

library(dplyr)
# Change value and TextualResults from factor to character
df_test[-1] <- lapply(df_test[-1], as.character)


df_test <- 
  df_test %>% 
  mutate(value = if_else(grepl("Index", Test, ignore.case = TRUE) & is.na(value), TextualResults, value))

df_test
            Test    value TextualResults
1         Sodium       12      low value
2    Heamo Index No haemo           <NA>
3    Lipae Index  NO LIPA        NO LIPA
4  ictoric index   No ict           <NA>
5       Chloride       21           HIGH
6 Blood pressure       32            low
7     Test Index  NO TEST        NO TEST

Answer 4

df_test %>%
   mutate(value = ifelse(is.na(value) & Test %like% 'index|Index|INDEX', as.character(TextualResults), as.character(value)))
            Test    value TextualResults
1         Sodium       12      low value
2   Heamo Index. No haemo           <NA>
3    Lipae Index  NO LIPA        NO LIPA
4  ictoric index   No ict           <NA>
5       Chloride       21           HIGH
6 Blood pressure       32            low
7     Test Index  NO TEST        NO TEST

让我们稍微修改一下：

Test <- c('Sodium','Heamo ..Index..','Lipae .Index_','ictoric ?index  ','Chloride','Blood pressure','Test Index')
value <- c(12,'No haemo',NA,'No ict',21,32,NA)
TextualResults <- c('low value',NA,'NO LIPA',NA,'HIGH','low','NO TEST')
df_test <- data.frame(Test,value,TextualResults)

df_test
              Test    value TextualResults
1           Sodium       12      low value
2  Heamo ..Index.. No haemo           <NA>
3    Lipae .Index_     <NA>        NO LIPA
4 ictoric ?index     No ict           <NA>
5         Chloride       21           HIGH
6   Blood pressure       32            low
7       Test Index     <NA>        NO TEST

结果是一样的：

              Test    value TextualResults
1           Sodium       12      low value
2  Heamo ..Index.. No haemo           <NA>
3    Lipae .Index_  NO LIPA        NO LIPA
4 ictoric ?index     No ict           <NA>
5         Chloride       21           HIGH
6   Blood pressure       32            low
7       Test Index  NO TEST        NO TEST