基于 r 中 ids 内的值对数据进行子集化

Question

我正在尝试根据两个标准对数据集进行子集化。这是我的数据快照：

ids <-  c(1,1,1,1,1,1, 2,2,2,2,2,2, 3,3,3,3,3,3)
seq <-  c(1,2,3,4,5,6, 1,2,3,4,5,6, 1,2,3,4,5,6)
type <- c(1,1,5,1,1,1, 1,1,1,8,1,1, 1,1,1,1,1,1)
    data <- data.frame(ids, seq, type)

   ids seq type
1    1   1    1
2    1   2    1
3    1   3    5
4    1   4    1
5    1   5    1
6    1   6    1
7    2   1    1
8    2   2    1
9    2   3    1
10   2   4    8
11   2   5    1
12   2   6    1
13   3   1    1
14   3   2    1
15   3   3    1
16   3   4    1
17   3   5    1
18   3   6    1

ids 是学生 ID，seq 是学生回答的问题（项目）的顺序。 type 指问题的类型。 1 是简单的，5 或 8 是复杂的项目。我想做的是生成第一个变量（complex）来判断学生是否有复杂的项目（type=5|8）。然后我想得到：

   > data
   ids seq type complex
1    1   1    1       1
2    1   2    1       1
3    1   3    5       1
4    1   4    1       1
5    1   5    1       1
6    1   6    1       1
7    2   1    1       1
8    2   2    1       1
9    2   3    1       1
10   2   4    8       1
11   2   5    1       1
12   2   6    1       1
13   3   1    1       0
14   3   2    1       0
15   3   3    1       0
16   3   4    1       0
17   3   5    1       0
18   3   6    1       0

第二步是在学生内部拆分数据。 (a) 对于有非复杂项目的学生（complex=0），我想将数据集从半点拆分并得到以下结果：

>simple.split.1
    ids seq type   complex
13   3   1    1       0
14   3   2    1       0
15   3   3    1       0

>simple.split.2
    ids seq type   complex
16   3   4    1       0
17   3   5    1       0
18   3   6    1       0

(b)对于有复杂项目的学生（complex=1），我想将复杂项目设置为切割点并从那里分割数据。所以数据应该是这样的（不包括复杂项）：

   >complex.split.1
    ids seq type   complex
1    1   1    1       1
2    1   2    1       1
7    2   1    1       1
8    2   2    1       1
9    2   3    1       1

    >complex.split.2
    ids seq type   complex
4    1   4    1       1
5    1   5    1       1
6    1   6    1       1
11   2   5    1       1
12   2   6    1       1

有什么想法吗？ 谢谢

Answer 1

这是一种使用data.table、zoo 包和split 函数的方法：

library(data.table)
library(zoo)

setDT(data)[, complex := ifelse(type == 5 | type == 8, 1, NA_integer_), by = ids][, complex := na.locf(na.locf(complex, na.rm=FALSE), na.rm=FALSE, fromLast=TRUE), by = ids][, complex := ifelse(is.na(complex), 0, complex)] ## set data to data.table & add a flag 1 where type is 5 or 8 ## carry forward and backward of complex flag ## replace na values in complex column with 0

data <- data[!(type == 5 | type == 8), ] ## removing rows where type equals 5 or 8

complex <- split(data, data$complex) ## split data based on complex flag

complex_0 <- as.data.frame(complex$`0`) ## saving as data frame based on complex flag
complex_1 <- as.data.frame(complex$`1`)

split(complex_0, cut(complex_0$seq, 2)) ## split into equal parts
split(complex_1, cut(complex_1$seq, 2))


#$`(0.995,3.5]`
#   ids seq type complex
#1   3   1    1       0
#2   3   2    1       0
#3   3   3    1       0

#$`(3.5,6]`
#   ids seq type complex
#4   3   4    1       0
#5   3   5    1       0
#6   3   6    1       0



    #$`(0.995,3.5]`
#   ids seq type complex
#1   1   1    1       1
#2   1   2    1       1
#6   2   1    1       1
#7   2   2    1       1
#8   2   3    1       1

#$`(3.5,6]`
#    ids seq type complex
#3    1   4    1       1
#4    1   5    1       1
#5    1   6    1       1
#9    2   5    1       1
#10   2   6    1       1

Answer 2

如果您更喜欢使用tidyverse，这里有一个方法：

ids <-  c(1,1,1,1,1,1, 2,2,2,2,2,2, 3,3,3,3,3,3)
seq <-  c(1,2,3,4,5,6, 1,2,3,4,5,6, 1,2,3,4,5,6)
type <- c(1,1,5,1,1,1, 1,1,1,8,1,1, 1,1,1,1,1,1)
data <- data.frame(ids, seq, type)

step1.data <- data %>%
  group_by(ids) %>%
  mutate(complex = ifelse(any(type %in% c(5,8)), 1, 0)) %>%
  ungroup()

simple.split.1 <- step1.data %>%
  filter(complex == 0) %>%
  group_by(ids) %>%
  filter(seq <= mean(seq)) %>% #if you happen to have more than 6 questions in seq, this gives the midpoint
  ungroup()

simple.split.2 <- step1.data %>%
  filter(complex == 0) %>%
  group_by(ids) %>%
  filter(seq > mean(seq)) %>%
  ungroup()

complex.split.1 <- step1.data %>%
  filter(complex == 1) %>%
  arrange(ids, seq) %>%
  group_by(ids) %>%
  filter(seq < min(seq[type %in% c(5,8)])) %>%
  ungroup()

complex.split.2 <- step1.data %>%
  filter(complex == 1) %>%
  arrange(ids, seq) %>%
  group_by(ids) %>%
  filter(seq > min(seq[type %in% c(5,8)])) %>%
  ungroup()