r - 在ggplot2中按多边形裁剪stat_density2d图

Question

我正在绘制美国大陆 github 配置文件的 2d 密度图。数据框如下所示：

> head(f)
    login id        location      lat       long
1 defunkt  2   San Francisco 37.78300 -122.41700
2 pjhyett  3   San Francisco 37.78300 -122.41700
3  wycats  4   San Francisco 37.78300 -122.41700
4    ivey  6 Bay Minette, AL 30.88333  -87.77694
5 evanphx  7 Los Angeles, CA 34.05000 -118.25000
6 vanpelt 17   San Francisco 37.78300 -122.41700

到目前为止，我想出了这段代码：

library(maps)
library(mapdata)
library(ggplot2)

all_states <- map_data("state")
p <- ggplot() + 
  geom_polygon( data=all_states, aes(x=long, y=lat, group = group),color="grey60", fill="white") +
  stat_density2d(data=f, aes(x=long, y=lat,fill = ..level..),geom='polygon') +
  theme(
    axis.line = element_blank(), 
    axis.text.x = element_blank(), 
    axis.text.y = element_blank(),
    axis.ticks = element_blank(), 
    axis.title.x = element_blank(), 
    axis.title.y = element_blank(),
    panel.grid.major = element_blank(),
    panel.grid.minor = element_blank(),
    panel.border = element_blank(),
    panel.background = element_blank()
  ) 

print(p)

此代码生成以下图像：

如何剪裁美国边境上方的密度区域？