总有办法的。

这里有一个解决方案，允许完全任意轴而不重新缩放。其思想是生成两个除了轴以外完全相同的图，并使用cowplot包中的insert_yaxis_grob和get_y_axis函数将它们组合在一起。

library(ggplot2)
library(cowplot)

## first plot 
p1 <- ggplot(mtcars,aes(disp,hp,color=as.factor(am))) + 
    geom_point() + theme_bw() + theme(legend.position='top', text=element_text(size=16)) +
    ylab("Horse points" )+ xlab("Display size") + scale_color_discrete(name='Transmitter') +
    stat_smooth(se=F)

## same plot with different, arbitrary scale   
p2 <- p1 +
    scale_y_continuous(position='right',breaks=seq(120,173,length.out = 3),
                       labels=c('little','medium little','medium hefty'))

ggdraw(insert_yaxis_grob(p1,get_y_axis(p2,position='right')))

以下内容结合了Dag Hjermann的基本数据和编程，改进了user4786271创建“转换函数”的策略，以优化组合图和数据轴，并响应了浸信会的提示，这样的函数可以在R中创建。

#Climatogram for Oslo (1961-1990)
climate <- tibble(
  Month = 1:12,
  Temp = c(-4,-4,0,5,11,15,16,15,11,6,1,-3),
  Precip = c(49,36,47,41,53,65,81,89,90,84,73,55))

#y1 identifies the position, relative to the y1 axis, 
#the locations of the minimum and maximum of the y2 graph.
#Usually this will be the min and max of y1.
#y1<-(c(max(climate$Precip), 0))
#y1<-(c(150, 55))
y1<-(c(max(climate$Precip), min(climate$Precip)))

#y2 is the Minimum and maximum of the secondary axis data.
y2<-(c(max(climate$Temp), min(climate$Temp)))

#axis combines y1 and y2 into a dataframe used for regressions.
axis<-cbind(y1,y2)
axis<-data.frame(axis)

#Regression of Temperature to Precipitation:
T2P<-lm(formula = y1 ~ y2, data = axis)
T2P_summary <- summary(lm(formula = y1 ~ y2, data = axis))
T2P_summary   

#Identifies the intercept and slope of regressing Temperature to Precipitation:
T2PInt<-T2P_summary$coefficients[1, 1] 
T2PSlope<-T2P_summary$coefficients[2, 1] 


#Regression of Precipitation to Temperature:
P2T<-lm(formula = y2 ~ y1, data = axis)
P2T_summary <- summary(lm(formula = y2 ~ y1, data = axis))
P2T_summary   

#Identifies the intercept and slope of regressing Precipitation to Temperature:
P2TInt<-P2T_summary$coefficients[1, 1] 
P2TSlope<-P2T_summary$coefficients[2, 1] 


#Create Plot:
ggplot(climate, aes(Month, Precip)) +
  geom_col() +
  geom_line(aes(y = T2PSlope*Temp + T2PInt), color = "red") +
  scale_y_continuous("Precipitation", sec.axis = sec_axis(~.*P2TSlope + P2TInt, name = "Temperature")) +
  scale_x_continuous("Month", breaks = 1:12) +
  theme(axis.line.y.right = element_line(color = "red"), 
        axis.ticks.y.right = element_line(color = "red"),
        axis.text.y.right = element_text(color = "red"), 
        axis.title.y.right = element_text(color = "red")) +
  ggtitle("Climatogram for Oslo (1961-1990)")

Most noteworthy is that a new "transformation function" works better with just two data points from the data set of each axes—usually the maximum and minimum values of each set. The resulting slopes and intercepts of the two regressions enable ggplot2 to exactly pair the plots of the minimums and maximums of each axis. As user4786271 pointed out, the two regressions transform each data set and plot to the other. One transforms the break points of the first y axis to the values of the second y axis. The second transforms the data of the secondary y axis to be "normalized" according to the first y axis. The following output shows how the axis align the minimums and maximums of each dataset:

使最大值和最小值匹配可能是最合适的;但是，这种方法的另一个好处是，如果需要，可以通过更改与主轴数据相关的编程行轻松地移动与次要轴相关的绘图。下面的输出只是将y1编程行中输入的最小降水量更改为“0”，从而将最小温度水平与“0”降水水平对齐。

从:y1<-(c(max(气候$ precp)， min(气候$ precp)))

到:y1<-(c(max(气候$ precp)， 0))

请注意，生成的新回归和ggplot2如何自动调整绘图和轴，以正确地将最低温度与“0”降水水平的新“基数”对齐。同样，可以很容易地提升Temperature图，使其更加明显。下面的图是通过简单地将上面提到的线更改为:

“日元<——(c(150年,55岁))”

上面的线表示温度曲线的最大值与“150”降水水平相吻合，温度曲线的最小值与“55”降水水平相吻合。再次注意，ggplot2和由此产生的新的回归输出如何使图保持与轴的正确对齐。

以上可能不是理想的输出;然而，这是一个例子，说明了如何容易地操纵图形，并且在图和轴之间仍然有正确的关系。 Dag Hjermann的主题的结合提高了与情节对应的轴的识别。

我承认并同意哈德利(和其他人)的观点，即单独的y量表“存在根本缺陷”。说到这里，我经常希望ggplot2有这个特性——特别是当数据是宽格式的，并且我想快速地可视化或检查数据时(即仅供个人使用)。

虽然tidyverse库可以很容易地将数据转换为长格式(这样facet_grid()就可以工作)，但这个过程仍然不是简单的，如下所示:

library(tidyverse)
df.wide %>%
    # Select only the columns you need for the plot.
    select(date, column1, column2, column3) %>%
    # Create an id column – needed in the `gather()` function.
    mutate(id = n()) %>%
    # The `gather()` function converts to long-format. 
    # In which the `type` column will contain three factors (column1, column2, column3),
    # and the `value` column will contain the respective values.
    # All the while we retain the `id` and `date` columns.
    gather(type, value, -id, -date) %>%
    # Create the plot according to your specifications
    ggplot(aes(x = date, y = value)) +
        geom_line() +
        # Create a panel for each `type` (ie. column1, column2, column3).
        # If the types have different scales, you can use the `scales="free"` option.
        facet_grid(type~., scales = "free")

这在ggplot2中是不可能的，因为我认为具有单独y尺度的图(不是相互转换的y尺度)从根本上是有缺陷的。一些问题:

The are not invertible: given a point on the plot space, you can not uniquely map it back to a point in the data space. They are relatively hard to read correctly compared to other options. See A Study on Dual-Scale Data Charts by Petra Isenberg, Anastasia Bezerianos, Pierre Dragicevic, and Jean-Daniel Fekete for details. They are easily manipulated to mislead: there is no unique way to specify the relative scales of the axes, leaving them open to manipulation. Two examples from the Junkcharts blog: one, two They are arbitrary: why have only 2 scales, not 3, 4 or ten?

你也可能想要阅读Stephen Few关于双缩放轴在图形中的主题的冗长讨论，它们是最好的解决方案吗?

总有办法的。

这里有一个解决方案，允许完全任意轴而不重新缩放。其思想是生成两个除了轴以外完全相同的图，并使用cowplot包中的insert_yaxis_grob和get_y_axis函数将它们组合在一起。

library(ggplot2)
library(cowplot)

## first plot 
p1 <- ggplot(mtcars,aes(disp,hp,color=as.factor(am))) + 
    geom_point() + theme_bw() + theme(legend.position='top', text=element_text(size=16)) +
    ylab("Horse points" )+ xlab("Display size") + scale_color_discrete(name='Transmitter') +
    stat_smooth(se=F)

## same plot with different, arbitrary scale   
p2 <- p1 +
    scale_y_continuous(position='right',breaks=seq(120,173,length.out = 3),
                       labels=c('little','medium little','medium hefty'))

ggdraw(insert_yaxis_grob(p1,get_y_axis(p2,position='right')))

根据上面的答案和一些微调(无论它有什么价值)，这里有一种通过sec_axis实现两个尺度的方法:

假设有一个简单的(完全虚构的)数据集dt:在五天的时间里，它追踪了被打断的次数VS工作效率:

        when numinter prod
1 2018-03-20        1 0.95
2 2018-03-21        5 0.50
3 2018-03-23        4 0.70
4 2018-03-24        3 0.75
5 2018-03-25        4 0.60

(两列的范围相差大约5倍)。

下面的代码将画出它们占用整个y轴的两个级数:

ggplot() + 
  geom_bar(mapping = aes(x = dt$when, y = dt$numinter), stat = "identity", fill = "grey") +
  geom_line(mapping = aes(x = dt$when, y = dt$prod*5), size = 2, color = "blue") + 
  scale_x_date(name = "Day", labels = NULL) +
  scale_y_continuous(name = "Interruptions/day", 
    sec.axis = sec_axis(~./5, name = "Productivity % of best", 
      labels = function(b) { paste0(round(b * 100, 0), "%")})) + 
  theme(
      axis.title.y = element_text(color = "grey"),
      axis.title.y.right = element_text(color = "blue"))

下面是结果(上面的代码+一些颜色调整):

重点(除了在指定y_scale时使用sec_axis之外)是在指定系列时将第二个数据系列的每个值与5相乘。为了在sec_axis定义中获得正确的标签，它需要除以5(并格式化)。因此，上述代码中的关键部分实际上是geom_line和~中的*5。sec_axis中的/5(一个除当前值的公式。5)。

相比之下(我不想在这里判断方法)，这是两个图表叠加在一起的样子:

你可以自己判断哪一个能更好地传递信息(“不要打扰别人工作!”)。我想这是一个公平的决定方式。

这两个图像的完整代码(实际上并没有比上面更多，只是完成并准备运行)在这里:https://gist.github.com/sebastianrothbucher/de847063f32fdff02c83b75f59c36a7d更详细的解释在这里:https://sebastianrothbucher.github.io/datascience/r/visualization/ggplot/2018/03/24/two-scales-ggplot-r.html