r - 以尺寸图例 geom_point ggplot2 显示所有形状

Question

我的图中使用了 3 个形状，size legend 中仅使用了实心圆(如图所示)。我试图找到一种方法来包含该图例中的所有形状，有没有办法做到这一点？

数据框和绘图代码:

data <- structure(list(date = structure(c(1431388800, 1444780800, 1456876800, 
    1469145600, 1469664000, 1425081600, 1445299200, 1488758400, 1524960000, 
    1454544000, 1540512000, 1429228800, 1481587200, 1526688000, 1445904000, 
    1498348800, 1464825600, 1451174400, 1450310400, 1539216000, 1528934400, 
    1525996800, 1465084800, 1531180800), class = c("POSIXct", "POSIXt"
    ), tzone = "UTC"), cnt1 = c(76, 140, 99, 10, 123, 124, 160, 58, 
    47, 43, 86, 112, 26, 123, 111, 49, 65, 93, 174, 49, 5, 16, 84, 
    4), cnt2 = c(111, 200, 550, 73, 123, 184, 166, 131, 189, 136, 
    250, 195, 26, 154, 125, 100, 168, 151, 255, 49, 200, 144, 177, 
    4), per = c(68.4684684684685, 70, 18, 13.6986301369863, 100, 
    67.3913043478261, 96.3855421686747, 44.2748091603053, 24.8677248677249, 
    31.6176470588235, 34.4, 57.4358974358974, 100, 79.8701298701299, 
    88.8, 49, 38.6904761904762, 61.5894039735099, 68.2352941176471, 
    100, 2.5, 11.1111111111111, 47.4576271186441, 100), status = c("A", 
    "B", "C", "C", "B", "B", "B", "A", "A", "A", "C", "C", "C", "A", 
    "B", "C", "B", "C", "A", "A", "A", "B", "B", "B"), category = c("CAT3", 
    "CAT1", "CAT3", "CAT2", "CAT1", "CAT1", "CAT1", "CAT3", "CAT2", 
    "CAT3", "CAT2", "CAT1", "CAT1", "CAT1", "CAT2", "CAT2", "CAT1", 
    "CAT1", "CAT1", "CAT1", "CAT3", "CAT3", "CAT3", "CAT1")), .Names = c("date", 
    "cnt1", "cnt2", "per", "status", "category"), row.names = c(NA, 
    -24L), class = c("tbl_df", "tbl", "data.frame"))



 ggplot(data, aes(x=date, y=cnt2)) + 
      scale_y_log10(breaks=c(1,10,100,1000,5000)) +
      scale_size_continuous(trans="sqrt",range = c(1, 5),breaks=c(5,10,25,50,75,100)) +
      geom_point(aes(col=category, size=per, shape=status), alpha=0.7)  

Answer 1

我找到了一种解决方法，因为似乎没有简单的方法可以做到这一点。理想的解决方案是让形状彼此相邻并在图例中提及一次数字而不是重复它。此外，由于 status 的数量不断变化，因此此解决方法可以更加参数化。并显示一个新的形状，所有输入到覆盖函数的向量都可以正确构造:

    l=c(5,10,25,50,75,100)
    mybreaks=c(rep(l,3))
    myshapes=c(rep(15,6),rep(16,6),rep(17,6))

    ggplot(data, aes(x=date, y=cnt2)) + 
    scale_y_log10(breaks=c(1,10,100,1000,5000)) +
    scale_size_continuous(trans="sqrt",range = c(1, 5),breaks=mybreaks) +
    geom_point(aes(col=category, size=per, shape=status), alpha=0.7)  +
    guides(size=guide_legend(ncol=3,override.aes = list( shape =myshapes)))