如何顯示具有較大價值差異的繪圖數據？

Question

我有我使用ggplot繪圖的數據car_crashes。它有3個不同的數據集所看到如下

但由於汽車的平均是巨大的，其它的值​​沒有表現出偶位，因爲它們是在100範圍內。如果我除去汽車數據的平均值，陰謀其實看起來是這樣的

有沒有一種方法，我可以顯示所有的數據在一個情節，以便至少我可以看到崩潰的數量情節？ 我使用的代碼是下面：

carcrashes_figure <- ggplot()+geom_area(aes(YEAR_WW,AverageofCars,group = 1,colour = 'Average of cars'),car_crashes,fill = "dodgerblue1",alpha = 0.4)+ 
    geom_line(aes(YEAR_WW,averageofcars,group = 1,linetype ='num of crashes'),car_crashes,fill = "dodgerblue3",colour = "dodgerblue3",size = 1.6) + 
    geom_line(aes(car_crashes$YEAR_WW,constantline,group = 1, size = 'constant line'),car_crashes1,fill = "green4",colour = "green4")+ 
    theme_bw() + 
    theme(axis.text.x = element_text(angle=70, vjust=0.6, face = 'bold'))+ 
    theme(axis.text.y = element_text(angle=0, vjust=0.2, face = 'bold'))+ 
    scale_colour_manual('', values = "dodgerblue1")+ 
    scale_size_manual('',values = 1.4)+ 
    scale_linetype_manual('',values = 1)+ 
    scale_y_continuous()+ 
    theme(legend.text = element_text(size = 8, colour = "black", angle = 0)) 
carcrashes_figure 

Answer 1

我同意的想法，使用單獨的y軸由@Jim夸克。據我所知，ggplot2做得不是很好，所以我用了基本圖。

# making example ts_data 
set.seed(1); data <- matrix(c(rnorm(21, 1000, 100), rnorm(21, 53, 10), rep(53, 21)), ncol=3) 
ts_data <- ts(data, start = 1980, frequency = 1) 

par(mar=c(4, 4.2, 1.5, 4.2))    # enlarge a right margin 
# plot(ts_data[,1])      # check y-range 
plot(ts_data[,2:3], plot.type = "single", ylab="num of crashes & constant line", 
    col=c(2,3), ylim=c(35,100), lwd=2) # draw "num of crashes" and "constant line" 
par(usr = c(par("usr")[1:2], 490, 1310)) # set the second y coordinates 
axis(4)         # write it on the right side 
polygon(x = c(1980:2000, rev(1980:2000)), y = c(ts_data[,1], rep(0,21)), 
     col="#0000FF20", border = "blue") # paint "Average of cars" 
mtext(side=4, "Average of cars", line=2.5) 
legend("topright",paste(c("num of crashes","constant line","Average of cars")), 
    pt.cex=c(0,0,3), lty=c(1,1,0), pch=15, cex=0.9, col=c(2, 3, "#0000FF20"), bty="n", 
    inset=c(0.02,-0.02), y.intersp=1.5)