如何在条形图中扩展geom_bspline？

Question

我研究了19世纪几十年的物种灭绝。我正在绘制一个光滑的样条，它显示了灭绝物种的累积值，横跨一组显示每十年绝对值的几条。现在，光滑的样条结束在最后一条的中心，因为值发生在X位置的中心。由于这在视觉上令人不快，我想扩展样条，以便它继续边界的情节。

我尝试了position的论点，但躲闪没有帮助。geom_bspline函数似乎也没有提供解决方案(例如，type="open“没有帮助)。

因此，我正在寻找一个解决方案，允许从ggforce包(或类似的解决方案)扩展geom_bspline。

下面我添加了绘图和绘图的代码示例。

​

​

library("tidyverse")
library("ggforce")


## DATASET
dftest <- tibble(
  x = seq(1,11,1),
  y1 = c(3,   0 ,  1  , 0 ,  2   ,0,  20, 126, 426, 400 , 37),
  y2 = c( 3 ,3   , 4 ,   4 ,   6 ,   6 ,  26,  152 , 578 , 978, 1015)
)


###### Figure ------
ggplot(dftest, aes(x=x)) + 
  
  ### BARS
  geom_bar(aes(y= y1), fill="#FC301A", stat="identity") +
  
  ### SPLINE
  geom_bspline(aes(x=x,y=y2),data=dftest)+
  
  ### SCALES
  scale_x_discrete(name ="",  drop = FALSE) + 
  scale_y_continuous(expand = c(0,0),
                     limits = c(0, 1200),
                     breaks = seq(0,1200,200),name = "Kumulativ",
                     sec.axis = sec_axis(~ ((. - a)/b)+2,
                                         name = "Absolut"),
                     labels=scales::comma_format(big.mark = ".")) +
  
  ### THEME
  theme_classic()+
  theme(
    panel.grid.major.y = element_line(colour = "grey70", size = 0.2),  # linien
    
    ###axis
    axis.ticks.length = unit(1, "mm"),
    
    axis.text.x = element_text(family = "Helvetica",  size=10,angle = 90,vjust = 0.5,
                               margin = unit(c(0.2,0,0,0),"cm")),
    axis.text.y = element_text(family = "Helvetica",  size=10),
    axis.text.y.right = element_text(color = "red",family = "Helvetica",  size=10),
    
    
    axis.title.y.right =  element_text(color = "red",family = "Helvetica",  size=10,
                                       margin = unit(c(0,0,0,0.25),"cm")),
    axis.title.y.left =  element_text(family = "Helvetica",  size=10, 
                                      margin = unit(c(0,0.25,0,0),"cm"))
    
  )

Answer 1

样条必须在节点之间进行插值，并且除非外推器是端节点，否则不能外推，因此，最简单的方法是创建数据帧的副本，并在x= 11.5或x= 11.7 (只要您认为“看起来对”的地方)添加额外的行和额外的时间度量，但是具有相同的y2值：

dftest2 <- rbind(dftest, tail(dftest, 1))
dftest2$x[nrow(dftest2)] <- 11.7

然后你就可以：

ggplot(dftest, aes(x = x)) + 
  geom_bar(aes(y= y1), fill="#FC301A", stat="identity") +
  geom_bspline(aes(x = x, y = y2), data = dftest2) +
  scale_x_discrete(name ="",  drop = FALSE) + 
  scale_y_continuous(expand = c(0,0),
                     limits = c(0, 1200),
                     breaks = seq(0,1200,200),name = "Kumulativ",
                     sec.axis = sec_axis(~ ((. - 0)/2.3)+2,
                                         name = "Absolut"),
                     labels=scales::comma_format(big.mark = ".")) +
  theme_classic()+
  theme(
    panel.grid.major.y = element_line(colour = "grey70", size = 0.2),
    axis.ticks.length = unit(1, "mm"),
    
    axis.text.x = element_text(family = "Helvetica",  
                               size=10,angle = 90, vjust = 0.5,
                               margin = unit(c(0.2,0,0,0),"cm")),
    axis.text.y = element_text(family = "Helvetica",size=10),
    axis.text.y.right = element_text(color = "red",
                                     family = "Helvetica",  size = 10),
    axis.title.y.right =  element_text(color = "red",family = "Helvetica",  
                                       size=10,
                                       margin = unit(c(0,0,0,0.25),"cm")),
    axis.title.y.left =  element_text(family = "Helvetica",  size=10, 
                                      margin = unit(c(0,0.25,0,0),"cm"))
  )

​

​

请注意，您的代码并不是完全可复制的，因为您的次轴没有包含变量a和b，所以我不得不在这里近似您的第二轴。