在 R 中合并的"special"案例

Question

我是一个相当缺乏经验的R用户，面临以下问题:

我想合并两个数据表dt1和dt2。dt1包含一个名为Assessment的变量。dt2包含两个名为ID和Frequency的变量。现在，我还想在dt2中获得评估观察结果。

为简单起见，考虑以下示例:

library(dplyr)
library(data.table)
dt1 <- data.table(c("perfect", "perfect", "okay", "unsufficient", "good", "good", "okay", "perfect"))
colnames(dt1) <- "Assessment"
dt2 <- data.table(cbind(c(1,2,3,4,5,6),c(1,3,1,1,1,1)))
colnames(dt2) <- c("ID", "Frequency")

因此，dt1看起来像:

tbody> <<tr>完美完美

评估
完美
好
unsufficient
好

dt1 %>% 
bind_cols(
dt2 %>% 
uncount(Frequency) 
) %>% 
group_by(ID) %>% 
summarise(Assessment = paste0(Assessment,collapse = ";"))
# A tibble: 6 x 2
ID Assessment               
<dbl> <chr>                    
1     1 perfect                  
2     2 perfect;okay;unsufficient
3     3 good                     
4     4 good                     
5     5 okay                     
6     6 perfect  

如果你相信正确的顺序，就像你在OP中说的，你可以根据它们的频率rep.intid。

dt2[dt1[, list(Assessment=toString(Assessment)), by=list(ID=with(dt2, rep.int(ID, Frequency)))], on=.(ID)]
#    ID Frequency                  Assessment
# 1:  1         1                     perfect
# 2:  2         3 perfect, okay, unsufficient
# 3:  3         1                        good
# 4:  4         1                        good
# 5:  5         1                        okay
# 6:  6         1                     perfect

或

dt2[dt1[, list(Assessment=list(Assessment)), by=list(ID=with(dt2, rep.int(ID, Frequency)))], on=.(ID)]
#    ID Frequency                Assessment
# 1:  1         1                   perfect
# 2:  2         3 perfect,okay,unsufficient
# 3:  3         1                      good
# 4:  4         1                      good
# 5:  5         1                      okay
# 6:  6         1                   perfect

不同的是，在第二个版本中，Assessment是一个列表列。

注意:如果dt2不包含任何其他内容，则无需再合并，并简化为

dt1[, list(Assessment=toString(Assessment)), by=list(ID=with(dt2, rep.int(ID, Frequency)))]
#    ID                  Assessment
# 1:  1                     perfect
# 2:  2 perfect, okay, unsufficient
# 3:  3                        good
# 4:  4                        good
# 5:  5                        okay
# 6:  6                     perfect