用两个y轴在R中绘图

所以我有一个数据框，看起来像这样：

我试图画出这个图，这样“癌症”在x轴上，然后“oddsRatio”和“foldIncrease”都是y轴参数。理想情况下，我有一个条形图，每个癌症的oddsRatio和FoldIngress相邻，并通过颜色进行区分

这是我能得到的最接近的结果：

使用以下代码：

cancers = c("All Cancers", "SCLC", "Lu SCC", "Lung AC", "Larynx", "Pharynx Oral", "Cavity", "Esophageal SCC", "Esophageal AC", "Bladder", "Liver", "Cervix", "Kidney", "Pancreas")
oddsRatio = c(1, 111.3, 103.5, 21.9, 13.2, 6.6, 4.2, 3.9, 3.9, 3.8, 2.9, 1.8, 1.7, 1.6)
foldIncrease = c(1.15464695360441, 0.858680717668245, 1.29986125563649, 4.56755645705811, 2.52922152922153, 0.818596228001855, 0.892133516869108, 1.04842410881178, 1.01791768793961, 1.1423932173087, 1.1871100629828, 0.857694671793762, 1.39573948596081, 1.33458517681876)

cancerData = data.frame(cancers, oddsRatio, foldIncrease)

par(mar = c(5,5,2,5))
with(cancerData, plot(cancers, oddsRatio, type="scatter", col="red3", 
             ylab='Odds Ratio',
             ylim=c(0,150)))

par(new = T)
with(cancerData, plot(cancers, foldIncrease, pch=16, axes=F, xlab=NA, ylab=NA, cex=1.2))
axis(side = 4)
mtext(side = 4, line = 3, 'Fold Increase')

为了澄清这一点，我正在寻找：

---

这就是你想要的吗？（注意，我使用了

ggplot2

而不是

base

）

库（tidyverse）
整洁的%
收集（统计数据，val，-癌症）
ggplot（tidy，aes（cancers，val））+
几何图形栏（aes（fill=stats），stat=“identity”，position=“dodge”）

编辑 好吧，一般来说我认为这不是个好主意。看看原因。但是如果你真的想，下面的内容可以帮你

vars <- length(unique(cancerData$cancer))

par(mar = c(4, 4, 4, 6))
plot(1:vars, seq(0, 120, length = vars), 
    type = "n",
    main = "Odds Ratios and Fold Increase",
    bty = "n",
    xaxt = "n",
    xlab = "Cancers",
    ylab = "Odds Ratios")
rect(1:vars - 0.25, 0, 1:vars, cancerData$oddsRatio, col = "blue")
axis(1, at = 1:vars, labels = unique(cancerData$cancer))

par(new = TRUE)
plot(1:vars, seq(0, 5, length = vars), 
    type = "n", 
    xaxt = "n", 
    yaxt = "n",
    bty = "n",
    xlab = "",
    ylab = "")
rect(1:vars, 0, 1:vars + 0.25, cancerData$foldIncrease, col = "red")
axis(4)
mtext("Fold Increase", side = 4, line = 3)
legend("topright", 
    fill = c("blue", "red"), 
    legend = c("Odds Ratio", "Fold Increase"),
    box.lwd = 0)

---

vars这就是你想要的吗？（注意，我使用了
ggplot2
而不是
base
）

库（tidyverse）
整洁的%
收集（统计数据，val，-癌症）
ggplot（tidy，aes（cancers，val））+
几何图形栏（aes（fill=stats），stat=“identity”，position=“dodge”）


编辑
好吧，一般来说我认为这不是个好主意。看看原因。但是如果你真的想，下面的内容可以帮你

vars <- length(unique(cancerData$cancer))

par(mar = c(4, 4, 4, 6))
plot(1:vars, seq(0, 120, length = vars), 
    type = "n",
    main = "Odds Ratios and Fold Increase",
    bty = "n",
    xaxt = "n",
    xlab = "Cancers",
    ylab = "Odds Ratios")
rect(1:vars - 0.25, 0, 1:vars, cancerData$oddsRatio, col = "blue")
axis(1, at = 1:vars, labels = unique(cancerData$cancer))

par(new = TRUE)
plot(1:vars, seq(0, 5, length = vars), 
    type = "n", 
    xaxt = "n", 
    yaxt = "n",
    bty = "n",
    xlab = "",
    ylab = "")
rect(1:vars, 0, 1:vars + 0.25, cancerData$foldIncrease, col = "red")
axis(4)
mtext("Fold Increase", side = 4, line = 3)
legend("topright", 
    fill = c("blue", "red"), 
    legend = c("Odds Ratio", "Fold Increase"),
    box.lwd = 0)

vars它很近，但最大的变化是我想要两个y轴，所以我想能够在它自己的轴上绘制褶皱增加，这样条就不会那么小。它很近，但最大的变化是我想要两个y轴，所以我想能够在它自己的轴上绘制褶皱增加，这样条就不会那么小了。