我有以下问题：我正在尝试使用权重（下限和上限）计算分类数据的 95% CI 范围。我有 0 或 1 的“响应”变量，我有 2 个因素（旧与新，以及 14 种不同颜色）。已收集每个参与者的权重。这是我的数据集：
dd <- data.frame(
weight = c(0.0037, 0.0016, 0.0347, 0.3421, 0.1047, 0.0065, 0.0153, 0.2856, 0.0032, 0.0321, 0.0321, 0.0321, 0.0321, 0.0321),
factor1 = factor(c("New", "Old", "New", "New", "New", "Old", "New", "Old", "New", "Old", "Old", "New", "Old", "Old")),
factor2 = factor(c("Red", "Yellow", "Green", "Orange", "Brown", "Blue", "Black", "White", "Purple", "Gray", "Pink", "Navy", "Tan", "Fuscia")),
question = factor(c("Q1", "Q2", "Q3", "Q4", "Q5", "Q6", "Q7", "Q8", "Q9", "Q10", "Q11", "Q12", "Q13", "Q14")),
response = c(0L, 1L, 1L, 1L, 1L, 1L, 0L, 1L, 0L, 1L, 1L, 0L, 1L, 0L)
)
我想为每个问题生成带有加权百分比的条形图，对于两个因素，如下所示：
question factor1 factor2 Weighted_pct Lower 95%CI Upper 95%CI
Q1 New Red 21.3% 21.1% 22.3%
Q1 Old Red 20.4% 20.1% 21%
Q1 New Red 22.6% 20.5% 22.3%
Q1 Old Red 19.6% 20.7% 22.3%
Q1 New Red 11.8% 20.9% 22.3%
Q1 Old Red 18.3% 20.4% 22.3%
Q1 New Red 27.4% 20.6% 22.3%
Q1 Old Red 6.3% 11% 22.3%
Q1 New Red 32.0% 20.3% 22.3%
Q1 Old Red 7.7% 20.3% 22.3%
Q1 New Red 9.3% 20.3% 22.3%
Q1 Old Red 15.3% 20.3% 22.3%
Q1 New Red 22.1% 20.3% 22.3%
Q1 Old Yellow 3.3% 20.3% 22.3%
数字是任意的，但我希望这是有道理的。我可以只使用一个因素让它工作，但是一旦我引入第二个因素，我的代码就会崩溃。
我当前的代码：
new_data <- complete_data %>%
dplyr::select(dplyr::Question, factor1, factor2, Response, weight) %>%
tidyr::gather(key = Question, value = Response, -weight, -factor1, -factor2) %>%
dplyr::group_by (Question, factor1, factor2) %>%
dplyr::summarise (Unweighted_N = n(),
Unweighted_pct = sum(Question) / Unweighted_N,
Effective_N = sum(weight),
Weighted_pct = sum (Question*weight)/Effective_N,
N_yes = round(Weighted_pct*Effective_N),
Low_95CI = binom.test(N_yes, round(Effective_N), conf.level = 0.95)$conf.int[1],
Up_96CI = binom.test(N=yes, round(Effective_N), conf.level = 0.95)$conf.int[2]) %>%
dplyr::select (Question,
factor 1,
factor 2,
Weighted_pct,
Low_95CI,
Up_95CI)