如何将 R 中嵌套的两组字符串中具有唯一值的行组合起来

Question

这有点复杂；我有一个数据框，其中有进行门诊就诊的患者，每次就诊都有药物，每列一种药物。在某些情况下，患者在一天内多次就诊，并且他们使用不同的药物 - 可能会记录更多、更少或相同数量的药物 - 每次就诊都会记录。发生这种情况时，我想将这些行合并为仅包含唯一药物的单行。下面，第一个数据帧是与我所拥有的类似的示例，第二个数据帧是所需的输出。

任何解决方案/指导将不胜感激。

    patient_id <- c(1, 1, 2, 2, 3, 3, 4, 4)
    dates <- c("2023-01-01", "2023-01-01", "2022-01-01", "2022-01-01", "2019-01-01", "2019-01-02", "2019-01-02", "2019-01-02")
    drug_1 <- c("paracetamol", NA, "aspirin", NA, NA,NA, NA, NA)
    drug_2 <- c("ibuprofen", "codine", NA, "paracetamol", NA, "asprin", NA, NA)
    drug_3 <- c("aspirin", "cough syrup", NA, NA, "ibuprofen", NA, NA, NA)
    drug_4 <- c(NA, "cocodamol", "ibuprofen", NA, NA, NA, NA, NA)
    drug_5 <- c(NA, "aspirin", NA, NA, NA, NA, NA, NA)
    
    patients_data <- data.frame(patient_id,dates, drug_1, drug_2, drug_3, drug_4, drug_5);patients_data

patient_id      dates      drug_1      drug_2      drug_3    drug_4  drug_5
1          1 2023-01-01 paracetamol   ibuprofen     aspirin      <NA>    <NA>
2          1 2023-01-01        <NA>      codine cough syrup cocodamol aspirin
3          2 2022-01-01     aspirin        <NA>        <NA> ibuprofen    <NA>
4          2 2022-01-01        <NA> paracetamol        <NA>      <NA>    <NA>
5          3 2019-01-01        <NA>        <NA>   ibuprofen      <NA>    <NA>
6          3 2019-01-02        <NA>      asprin        <NA>      <NA>    <NA>
7          4 2019-01-02        <NA>        <NA>        <NA>      <NA>    <NA>
8          4 2019-01-02        <NA>        <NA>        <NA>      <NA>    <NA>
    
    patient_id <- c(1, 2,3,3,4)
    dates <- c("2023-01-01", "2022-01-01", "2019-01-01", "2019-01-02", "2019-01-02")
    drug_1 <- c("paracetamol", "aspirin", NA, NA, NA)
    drug_2 <- c("ibuprofen", "ibuprofen", NA, "aspirin", NA)
    drug_3 <- c("aspirin", "paracetamol", "ibuprofen",NA, NA)
    drug_4 <- c("codine", NA, NA, NA, NA)
    drug_5 <- c("cough syrup", NA,NA, NA, NA)
    drug_6 <- c("cocodamol", NA,NA, NA, NA)
    
    patients_data <- data.frame(patient_id,dates, drug_1, drug_2, drug_3, drug_4, drug_5, drug_6);patients_data

  patient_id      dates      drug_1    drug_2      drug_3 drug_4      drug_5    drug_6
1          1 2023-01-01 paracetamol ibuprofen     aspirin codine cough syrup cocodamol
2          2 2022-01-01     aspirin ibuprofen paracetamol   <NA>        <NA>      <NA>
3          3 2019-01-01        <NA>      <NA>   ibuprofen   <NA>        <NA>      <NA>
4          3 2019-01-02        <NA>   aspirin        <NA>   <NA>        <NA>      <NA>
5          4 2019-01-02        <NA>      <NA>        <NA>   <NA>        <NA>      <NA>

Answer 1

你可以尝试这样的事情。

library(tidyverse)

patients_data %>%  
  pivot_longer(
    starts_with('drug')
  ) %>% 
  filter(
    !is.na(value)
  ) %>% 
  mutate(
    rn = row_number(),
    .by = c(patient_id, dates)
  ) %>% 
  mutate(
    name = str_glue('drug_{rn}')
  ) %>% 
  select(-rn) %>% 
  pivot_wider(
    names_from = name,
    values_from = value
  )
#> # A tibble: 4 x 9
#>   patient_id dates      drug_1      drug_2    drug_3 drug_4 drug_5 drug_6 drug_7
#>        <dbl> <chr>      <chr>       <chr>     <chr>  <chr>  <chr>  <chr>  <chr> 
#> 1          1 2023-01-01 paracetamol ibuprofen aspir~ codine cough~ cocod~ aspir~
#> 2          2 2022-01-01 aspirin     ibuprofen parac~ <NA>   <NA>   <NA>   <NA>  
#> 3          3 2019-01-01 ibuprofen   <NA>      <NA>   <NA>   <NA>   <NA>   <NA>  
#> 4          3 2019-01-02 asprin      <NA>      <NA>   <NA>   <NA>   <NA>   <NA>

^{创建于 2023-10-23，使用 reprex v2.0.2}