r - 如何将特定长度和数量的随机连续间隙插入向量中？

Question

我需要创建几个具有不同随机数和连续间隙长度的数据集。我想用 NA 创建一个向量来与原始数据集相乘。这是我目前为一个数据集获得的代码(整个模拟将使用 MonteCarlo 进行):

    For gap definition:

    size1 <- sample(1:40, size=1)
    length <- sample(1:480, size=size1)

    Expressing as different vectors with NA:      

    vec <- list()

    for (i in 1:length(length)){
    vec[[i]] <- seq(1:length[i])
    vec[[i]] <- replace(c(vec[[i]]), values=NA)
    }

我还需要将这些向量随机插入，但连贯且不相邻(间隔至少为 1)到特定长度(例如 37000)的向量中。

   x <- numeric(37000)
   x <- replace(x, 1)

有谁知道这样做的好方法吗？提前干杯!

Answer 1

我不确定这是最好的方法，但这是一个选择:

vector_length <- 37000L
num_gaps <- sample(40L, 1L) # 40 is the max amount of gaps
gap_lengths <- sample(480L, num_gaps) # 480 is the max length of a gap

# initialize
overlaps <- TRUE
gap_ends <- Inf
while (any(overlaps) || any(gap_ends > vector_length)) {
  gap_starts <- sort(sample(vector_length, num_gaps))
  gap_ends <- gap_starts + gap_lengths - 1L
  # make sure the distance between each gap is at least 1
  overlaps <- gap_starts[-1L] - gap_ends[-num_gaps] <= 1L
}

x <- numeric(vector_length)
for (i in seq_along(gap_starts)) {
  x[gap_starts[i]:gap_ends[i]] <- NA_real_
}