Question

我有两个具有不同主键的单独表，我需要将它们连接在一起。一个表提供结果，另一个表提供所涉及人员的ID。由于存在多个结果，并且可能有多个人具有相似的结果，因此我的表联接未正确同步。我可能在这里错过了一步，没有在逻辑上思考问题，但是任何建议都将不胜感激。

我最初尝试使用发现的此功能连接表：

rbind.all.columns <- function(x, y) {

    x.diff <- setdiff(colnames(x), colnames(y))
    y.diff <- setdiff(colnames(y), colnames(x))

    x[, c(as.character(y.diff))] <- NA

    y[, c(as.character(x.diff))] <- NA

    return(rbind(x, y))
}

但是，它所做的一切都为我提供了结果和ID列表。

表1：

EVENT   ID
145754  1738
145754  1756
145639  1738
145639  1756

df1 <- structure(list(EVENT = c(145754L, 145754L, 145639L, 145639L), ID = c(1738L, 1756L, 1738L, 1756L)), class = "data.frame", row.names = c(NA, -4L))

表2：

ENTRY   EVENT   RESULT
DEL     145754  Was given xxx med
INS     145754  
DEL     145639  Reported stomachache
INS     145639  

df2 <- structure(list(ENTRY = c("DEL", "INS", "DEL", "INS"), EVENT = c(145754L, 145754L, 145639L, 145639L), RESULT = c("Was given xxx med", "", "Reported stomachache", "")), class = "data.frame", row.names = c(NA, -4L))

所需表：

ID          EVENT   RESULT
1738, 1756  145754  Was given xxx med
1738, 1756  145639  Reported stomachache

Answer 1

我们需要做的是使用paste将同一事件的ID连接到一个逗号分隔的列表中：

library(tidyverse)
df1_concat <- df1 %>%
    group_by(EVENT) %>%
    summarise(IDs = paste(ID, collapse = ', '))

# A tibble: 2 x 2
   EVENT IDs       
   <int> <chr>     
1 145639 1738, 1756
2 145754 1738, 1756

然后，我们可以在“事件”列中进行*_join：

left_join(df2, df1_concat, by = 'EVENT')

  ENTRY  EVENT               RESULT        IDs
1   DEL 145754    Was given xxx med 1738, 1756
2   INS 145754                      1738, 1756
3   DEL 145639 Reported stomachache 1738, 1756
4   INS 145639                      1738, 1756

我不清楚为什么要删除ENTRY == 'INS'所在的行，但是根据此处的逻辑，有很多方法可以将其过滤掉。我在下面显示2：

# Remove rows where ENTRY == 'DEL'

left_join(df1_concat, df2, by = 'EVENT') %>%
    filter(ENTRY == 'DEL')

# A tibble: 2 x 4
   EVENT IDs        ENTRY RESULT              
   <int> <chr>      <fct> <fct>               
1 145639 1738, 1756 DEL   Reported stomachache
2 145754 1738, 1756 DEL   Was given xxx med   


# Remove rows with no value for RESULT

left_join(df1_concat, df2, by = 'EVENT') %>%
    filter(RESULT != '')

# A tibble: 2 x 4
   EVENT IDs        ENTRY RESULT              
   <int> <chr>      <fct> <fct>               
1 145639 1738, 1756 DEL   Reported stomachache
2 145754 1738, 1756 DEL   Was given xxx med

Answer 2

在基数R中，我们实际上可以使用aggregate，merge和toString（用于ID）的单线来完成此操作。

d <- aggregate(ID ~ EVENT + RESULT, merge(df1, df2[which(df2$RESULT != ""), ]), toString)
d
#    EVENT               RESULT         ID
# 1 145639 Reported stomachache 1738, 1756
# 2 145754    Was given xxx med 1738, 1756

是否有将多个主键信息组合到表中的功能？

2 个答案: