带有 geom_smooth 的视觉上不等间距的 log2 缩放轴

Question

我有一个数据集，其中包含按年龄（X 轴）递减的 HR 值（Y 轴），分为两组。

https://www.dropbox.com/s/l2p24llxcvndljl/reproducible_data_for_log2plot.txt?dl=0

我正在尝试创建两个 geom_smooth(method = "loess")，Y 轴 (HR) 采用 log2 比例，间距不均匀，受限于从 1 到 70 的指定中断。

读取数据：

aLotOfHRs <- read.table("https://www.dropbox.com/s/l2p24llxcvndljl/reproducible_data_for_log2plot.txt?dl=1" , header = TRUE , sep = "\t")

我的第一次尝试：

p <- ggplot(aLotOfHRs, aes(x = Age,y = HR,fill=Quantiles,color =Quantiles)) +
    geom_smooth(method = "loess" , formula = y~x) +
    # geom_point() +
    theme_minimal() +
    ylab("HR per 1-SD (log2 scale)") +
    xlim(c(40,72)) +
    scale_y_continuous(trans = scales::log2_trans())

1st Try

不幸的是，比例在视觉上是相等的，但 log2 图是正确的。如果我指定中断、标签和限制：

p <- ggplot(aLotOfHRs, aes(x = Age,y = HR,fill=Quantiles,color =Quantiles)) +
    geom_smooth(method = "loess" , formula = y~x) +
    # geom_point() +
    theme_minimal() +
    ylab("HR per 1-SD (log2 scale)") +
    xlim(c(40,72)) +
    scale_y_continuous(trans = scales::log2_trans()
        , breaks = log2(c(1.1,2,4,8,16,32,64))
        , labels = c(1.1,2,4,8,16,32,64)
        , limits = c(log2(1.1) , log2(70)))
# Note: I can't extend to 1 or it returns
# Error in seq.default(a, b, length.out = n + 1) :'from' must be a finite number

2nd Try

我得到了正确的比例但错误的情节。

迄今为止最好的结果是最后变换坐标，但我仍然无法按照我想要的方式设置中断：

p <- ggplot(aLotOfHRs, aes(x = Age,y = HR,fill=Quantiles,color =Quantiles)) +
    # geom_point() +
    theme_minimal() +
    ylab("HR per 1-SD of PRS (log2 scale)") +
    xlim(c(40,70)) +
    geom_smooth(method = "loess" , formula = y~x) +
    ylim(c(1,70) ) +
    coord_trans( y = "log2")

3d Try

有什么建议吗？

Answer 1

这可能是你所追求的：

aLotOfHRs <- read.table("https://www.dropbox.com/s/l2p24llxcvndljl/reproducible_data_for_log2plot.txt?dl=1" , header = TRUE , sep = "\t")
p <- ggplot(aLotOfHRs, aes(x = Age,y = HR,fill=Quantiles,color =Quantiles)) +
    # geom_point() +
    theme_minimal() +
    ylab("HR per 1-SD of PRS (log2 scale)") +
    xlim(c(40,70)) +
    geom_smooth(method = "loess" , formula = y~x)
p + scale_y_continuous(trans = scales::log2_trans(),breaks = c(1,2,4,8,16,32,64), lim = c(1,100), expand = c(0,0))

这是您的第三个电话，但最后添加了scale_y_continuous() 电话。