R geom_面积和坐标极坐标到plotly的转换

我正在尝试将ggplot极坐标图转换为plotly极坐标图。我在plotly中丢失了

geom_面积的累积值

，以及每个轨迹的填充颜色，而这两个轨迹都没有到达原点。目标是尽可能靠近ggplot视图

df <- data.frame(angle = rep(seq(0, 360, 45), 4), frequency = rep(1:9, 4), z = as.character(c(rep(1, 9), rep(2, 9))))
df <- aggregate(frequency ~ angle + z, data = df, sum)

pl <- ggplot(df,aes(x=angle,y=frequency,fill=z)) + geom_area() + coord_polar() + scale_x_continuous(breaks=angle, labels=c("N", "NE", "E", "SE", "S", "SO", "O", "NO", "N"))
pl

---

df以下是解决问题的起点：

k <- 100
df <- data.frame(angle = rep(seq(0, 360, length.out=k), 4), 
                 frequency = rep(1:k, 4), 
                 z = as.character(c(rep(1, k), rep(2, k))))
df <- aggregate(frequency ~ angle + z, data = df, sum)   
df$frequency[df$z==2] <- df$frequency[df$z==2]/2

p <- plot_ly(data=df, type = 'scatterpolar', mode = "lines") %>%
     add_trace(theta=~angle,r=~frequency,fill="toself", color=~z)
p <- p %>% layout(
  polar = list(radialaxis = list(angle = 90),
               angularaxis = list(
                 rotation = 90,
                 direction = 'clockwise',
                 ticktext =  c("N", "NE", "E", "SE", "S", "SO", "O", "NO", "N"),
                 tickvals = seq(0, 360, 45)
               )
  )
)
p

k非常感谢您抽出时间。事实上，图的平滑丢失了，你通过修改初始数据帧得到了它，这是我做不到的
k <- 100
df <- data.frame(angle = rep(seq(0, 360, length.out=k), 4), 
                 frequency = rep(1:k, 4), 
                 z = as.character(c(rep(1, k), rep(2, k))))
df <- aggregate(frequency ~ angle + z, data = df, sum)   
df$frequency[df$z==2] <- df$frequency[df$z==2]/2

p <- plot_ly(data=df, type = 'scatterpolar', mode = "lines") %>%
     add_trace(theta=~angle,r=~frequency,fill="toself", color=~z)
p <- p %>% layout(
  polar = list(radialaxis = list(angle = 90),
               angularaxis = list(
                 rotation = 90,
                 direction = 'clockwise',
                 ticktext =  c("N", "NE", "E", "SE", "S", "SO", "O", "NO", "N"),
                 tickvals = seq(0, 360, 45)
               )
  )
)
p