Ordering of traces is still wrong

Following on from #225, here is an an example where ordering of traces is wrong

set.seed(12345)

df1 = data.frame(
  x = seq(1, 200), y = 100 + cumsum(rnorm(200))
)

y.sparse = local({
  i = sample(200, 10)
  res = rep(0, 200)
  res[i] = 120 + cumsum(rnorm(10))
  res
})

df2 = data.frame(
  x = seq(1, 200), y = y.sparse
)

p = ggplot(mapping = aes(x = x, y = y)) +
  geom_area(data = df1) + 
  geom_col(data = df2, fill = "red")

p

ggplotly(p)

首先執行 issue 中的 R 重現，並比較原始 ggplot 輸出與 ggplotly(p) 的結果。追蹤 geom_area 和 geom_col 圖層如何轉換成 traces；當轉換後的 traces 保留來源圖中顯示的預期順序時，即表示完成。