如何根据r中的多个条件分组过滤案例?

Question

这是我的数据集:

df = data.frame(ID = c(1,1,1,2,2,2,3,3,3,4,4,4), 
Diagnosis=c("NEG","NEG","POS","NEG","NEG","NEG","NEG","POS","POS",'','',''))

我想用两个条件过滤掉重复的病例:1)如果"诊断"等于"POS"，那么选择一个"POS"观测值;2)如果"诊断"不等于"POS"，则检查"诊断"是否等于"NEG"，然后选择一个"NEG"观测值;3)对于不满足条件1和条件2的其他情况，只需选择组(ID)中的任意一条记录。

这是我想要的数据集:

df = data.frame(ID = c(1,2,3,4), 
Diagnosis=c("POS","NEG","POS",''))

这是我尝试的代码，但它没有得到预期的结果:

df_unique <- df %>% group_by(ID) %>% filter(Diagnosis==ifelse('POS' %in% Diagnosis, first(Diagnosis=='POS'),ifelse((!('POS' %in% Diagnosis)&('NEG' %in% Diagnosis)),first(Diagnosis=='POS'),first(Diagnosis))))

Answer 1

一种方法是将Diagnosis作为有序因子，然后将arrange(按此列排序)并按组选择第一行。

library(dplyr)
df$Diagnosis <- factor(df$Diagnosis, levels = c("POS", "NEG", ""), ordered = TRUE)
df %>%
group_by(ID) %>%
arrange(Diagnosis) %>%
slice(1)

ID Diagnosis
<dbl> <ord>    
1     1 "POS"    
2     2 "NEG"    
3     3 "POS"    
4     4 ""  

Answer 2

我们也可以使用distinct

library(dplyr)
library(forcats)
df %>%
mutate(Diagnosis = fct_relevel(Diagnosis, c("POS", "NEG", ""))) %>% 
arrange(ID, Diagnosis) %>% 
distinct(ID, .keep_all = TRUE)
#  ID Diagnosis
#1  1       POS
#2  2       NEG   
#3  3       POS
#4  4