新结构化变量的数据表解决方案

Question

data=data.frame("student"=c(1,1,1,1,2,2,2,2,3,3,4,4,4,4),
"score"=c(1,2,1,1,2,3,2,NA,3,NA,1,3,2,1),
"drop"=c(0,0,0,0,0,0,0,1,0,1,0,0,0,0),
"WANT"=c(1,2,1,1,2,3,3,4,3,4,1,3,3,3))

我有没有“WANT”的数据框“数据”，这是我希望使用 data.table 解决方案创建的。

规则是：

如果 score = 1, WANT = 1 如果 score = 2, WANT = 2 如果 score = 3, WANT = 3, 如果 drop = 1, WANT=4

如果在 t = 2 时得分并且在 t+1 = 1 时得分，那没关系，但是

如果 t = 3 的得分和任何以后的得分都小于 3，则将它们替换 3.

这意味着一个分数系列：1-2-1-3-1 should be: 1-2-1-3-3

    data2=data.frame("student"=c(1,1,1,1,2,2,2,2,3,3,4,4,4,4,5,5,5,5),
"score"=c(1,2,1,1,2,3,2,NA,3,NA,1,3,2,1,1,3,NA,2),
"drop"=c(0,0,0,0,0,0,0,1,0,1,0,0,0,0,0,0,0,0),
"WANT"=c(1,2,1,1,2,3,3,4,3,4,1,3,3,3,1,3,3,3))

Answer 1

我们可以在根据每个“学生”的“分数”中出现 3 个值创建条件后使用replace

library(dplyr)
data %>% 
   group_by(student) %>%
   mutate(WANT2 = replace(if(3 %in% score) replace(score, 
     (match(3, score) +1):n(), 3) else score, is.na(score) & drop == 1, 4))
# A tibble: 14 x 5
# Groups:   student [4]
#   student score  drop  WANT WANT2
#     <dbl> <dbl> <dbl> <dbl> <dbl>
# 1       1     1     0     1     1
# 2       1     2     0     2     2
# 3       1     1     0     1     1
# 4       1     1     0     1     1
# 5       2     2     0     2     2
# 6       2     3     0     3     3
# 7       2     2     0     3     3
# 8       2    NA     1     4     4
# 9       3     3     0     3     3
#10       3    NA     1     4     4
#11       4     1     0     1     1
#12       4     3     0     3     3
#13       4     2     0     3     3
#14       4     1     0     3     3

Answer 2

使用data.table的选项：

library(data.table)

#if score = 1, WANT = 1 if score = 2, WANT = 2 if score = 3, WANT = 3
setDT(data)[, w := score]

#if score at t = 3 and score at any later scores are less than 3, they are replaced with 3.
data[data[, .I[cummax(score)==3L & score < 3L], student]$V1, w := 3L]

#it add student '5' which has NA values that I hope to fill with prior non-missing NA value
data[, w := nafill(w, "locf")]

#if drop = 1, WANT=4
data[drop==1L, w := 4L]