如何根据值在总体中出现的速率 (R) 为观测值分配值

Question

我有一个数据框（df），其中有一列“问题”，其中详细说明了人们向急诊室提出的医疗状况（例如“扭伤”、“割伤”、“感染”）。

然而，有些人在被发现之前就离开了——在这种情况下，他们的问题被标记为“在被发现之前就离开了”。

我想更改我的数据框，以估算这些“在被看到之前留下”的人参加急诊室的问题是什么。

我有每个问题在人口中的比例：

sprain = 0.3
cut = 0.5
infection = 0.2.

如何根据这些问题在总体中出现的比率，将“扭伤”、“割伤”和“感染”随机分配给数据框中的“问题”=“被发现之前留下”的观察结果？这将出现在一个名为“impulated_problem”的新列中，这样我就不会丢失人们在被看到之前留下的数据。

注意：此数据是总体，而不是样本。

Answer 1

您可以使用

sample

将索引值随机分配给这三个问题的向量，并具有所需比例的给定计数。

n.left <- sum(df$problem=="left before being seen")

n.sprain <- round(n.left * 0.3)
n.cut <- round(n.left * 0.5)
n.infection <- n.left - (n.sprain + n.cut)

df$imputed_problem[df$problem=="left before being seen"] <- c(
  rep('sprain', n.sprain),
  rep('cut', n.cut), 
  rep('infection', n.infection))[sample(1:n.left)]

检查

prop.table(table(df$imputed_problem))

      cut infection    sprain 
      0.5       0.2       0.3

---

数据：

structure(list(problem = c("cut", "infection", "cut", "sprain", 
"cut", "cut", "sprain", "cut", "cut", "infection", "cut", "sprain", 
"sprain", "sprain", "infection", "sprain", "cut", "sprain", "infection", 
"sprain", "cut", "infection", "cut", "infection", "infection", 
"cut", "cut", "sprain", "infection", "sprain", "infection", "infection", 
"infection", "cut", "cut", "sprain", "infection", "sprain", "sprain", 
"cut", "infection", "sprain", "infection", "infection", "cut", 
"cut", "cut", "cut", "infection", "cut", "cut", "cut", "cut", 
"infection", "cut", "cut", "cut", "sprain", "sprain", "infection", 
"infection", "sprain", "sprain", "infection", "sprain", "sprain", 
"infection", "infection", "cut", "cut", "sprain", "cut", "infection", 
"infection", "cut", "cut", "cut", "cut", "cut", "infection", 
"cut", "sprain", "cut", "infection", "sprain", "sprain", "sprain", 
"cut", "cut", "cut", "infection", "cut", "sprain", "cut", "sprain", 
"cut", "sprain", "cut", "infection", "cut", "sprain", "cut", 
"cut", "cut", "cut", "cut", "cut", "cut", "sprain", "sprain", 
"sprain", "cut", "cut", "cut", "cut", "sprain", "sprain", "sprain", 
"sprain", "cut", "infection", "sprain", "infection", "sprain", 
"infection", "infection", "cut", "infection", "sprain", "cut", 
"infection", "infection", "infection", "sprain", "cut", "sprain", 
"cut", "sprain", "cut", "infection", "infection", "infection", 
"cut", "sprain", "infection", "sprain", "sprain", "cut", "sprain", 
"cut", "cut", "sprain", "sprain", "sprain", "cut", "cut", "cut", 
"cut", "cut", "sprain", "cut", "cut", "cut", "cut", "sprain", 
"sprain", "infection", "sprain", "cut", "cut", "cut", "sprain", 
"sprain", "cut", "cut", "cut", "infection", "sprain", "cut", 
"infection", "cut", "cut", "infection", "sprain", "sprain", "cut", 
"cut", "sprain", "cut", "sprain", "infection", "cut", "cut", 
"cut", "sprain", "cut", "cut", "cut", "infection", "cut", "cut", 
"infection", "cut", "sprain", "sprain", "infection", "sprain", 
"cut", "cut", "sprain", "infection", "infection", "sprain", "sprain", 
"cut", "sprain", "cut", "infection", "cut", "infection", "infection", 
"cut", "cut", "cut", "cut", "sprain", "cut", "infection", "cut", 
"cut", "sprain", "cut", "cut", "sprain", "cut", "cut", "cut", 
"infection", "cut", "sprain", "cut", "infection", "sprain", "sprain", 
"sprain", "cut", "cut", "sprain", "sprain", "cut", "cut", "sprain", 
"sprain", "cut", "infection", "sprain", "cut", "cut", "cut", 
"cut", "cut", "infection", "sprain", "sprain", "infection", "cut", 
"cut", "sprain", "cut", "sprain", "cut", "cut", "cut", "sprain", 
"sprain", "infection", "cut", "sprain", "cut", "sprain", "cut", 
"infection", "cut", "sprain", "cut", "cut", "infection", "infection", 
"cut", "cut", "sprain", "cut", "infection", "infection", "infection", 
"cut", "cut", "infection", "cut", "sprain", "sprain", "cut", 
"cut", "cut", "sprain", "infection", "cut", "cut", "cut", "cut", 
"cut", "sprain", "cut", "cut", "sprain", "cut", "infection", 
"cut", "sprain", "cut", "sprain", "infection", "sprain", "cut", 
"infection", "cut", "sprain", "infection", "sprain", "sprain", 
"cut", "infection", "sprain", "cut", "cut", "sprain", "sprain", 
"infection", "sprain", "cut", "sprain", "cut", "sprain", "cut", 
"cut", "infection", "cut", "cut", "cut", "sprain", "cut", "infection", 
"cut", "infection", "cut", "cut", "sprain", "infection", "infection", 
"cut", "cut", "sprain", "cut", "cut", "sprain", "infection", 
"cut", "cut", "infection", "sprain", "cut", "sprain", "cut", 
"cut", "cut", "sprain", "infection", "cut", "cut", "infection", 
"cut", "cut", "cut", "cut", "cut", "sprain", "sprain", "sprain", 
"infection", "cut", "sprain", "sprain", "sprain", "cut", "cut", 
"sprain", "cut", "cut", "cut", "cut", "cut", "cut", "sprain", 
"cut", "cut", "infection", "infection", "sprain", "infection", 
"cut", "infection", "cut", "cut", "cut", "sprain", "cut", "cut", 
"sprain", "sprain", "infection", "sprain", "cut", "cut", "sprain", 
"cut", "cut", "cut", "cut", "cut", "sprain", "infection", "infection", 
"sprain", "cut", "cut", "sprain", "infection", "sprain", "sprain", 
"cut", "cut", "cut", "cut", "cut", "sprain", "cut", "infection", 
"sprain", "sprain", "cut", "sprain", "cut", "cut", "sprain", 
"cut", "sprain", "cut", "cut", "cut", "sprain", "sprain", "cut", 
"sprain", "cut", "infection", "cut", "cut", "cut", "cut", "sprain", 
"cut", "cut", "cut", "cut", "sprain", "cut", "cut", "infection", 
"infection", "cut", "cut", "cut", "infection", "cut", "sprain", 
"cut", "cut", "sprain", "infection", "cut", "sprain", "sprain", 
"cut", "sprain", "infection", "infection", "cut", "cut", "cut", 
"sprain")), class = "data.frame", row.names = c(NA, -500L))