connect step function to 0 and max y-CodePudding

Using this code as example:

df <- data.frame(
        x = c(30, 40, 20, 60)
        , y = c(10, 20, 5, 70)
        , group = c("g1", "g1", "g2", "g2") 
    )

options(repr.plot.width = 10, repr.plot.height = 7) 
ggplot()  
    geom_point(data=df, aes(x=x, y=y, group=group, color=group))  
    geom_step(data=df, aes(x=x, y=y, group=group, color=group), size=1)  
    scale_y_continuous(breaks = round(seq(0, 100, by = 5), 1))  
    coord_cartesian(ylim = c(0, 100))

It produces:

Is there an easy way in ggplot (?) to achieve the following (only indicated by one step function):

Or do I need to change original data and somehow impute? Hope this makes sense? Thanks!

CodePudding user response：

You could add rows for each group's starting and ending points to match the desired result:

df <- df %>% 
  group_by(group) %>% 
  group_modify(~ add_row(.x, x=c(0, Inf), y=c(0, max(.x$y))))

#  group     x     y
#  <chr> <dbl> <dbl>
#1 g1       30    10
#2 g1       40    20
#3 g1        0     0
#4 g1      Inf    20
#5 g2       20     5
#6 g2       60    70
#7 g2        0     0
#8 g2      Inf    70

ggplot(df, aes(x=x, y=y, group=group, color=group))  
  geom_point()  
  geom_step(size=1)  
  scale_y_continuous(breaks = round(seq(0, 100, by = 5), 1))  
  coord_cartesian(ylim = c(0, 100), xlim = c(20,60))

Or, if you want this only for the first group (same idea; but not by group):

df %>% 
  add_row(., group = "g1", x = c(0, Inf), y = c(0, max(.$y[.$group == "g1"]))) %>% 
  ggplot(aes(x=x, y=y, group=group, color=group))  
  geom_point()  
  geom_step(size=1)  
  scale_y_continuous(breaks = round(seq(0, 100, by = 5), 1))  
  coord_cartesian(ylim = c(0, 100), xlim = c(20,60))

CodePudding user response：

Hope I get your question right.

You don't really need to "impute" anything, because what you want is to extend the geom_step() to 0 on the left hand side and Inf on the right hand side. So I added the corresponding values in df.

df <- data.frame(
  x = c(30, 40, 20, 60, Inf, -Inf),
  y = c(10, 20, 5, 70, 20, 0),
  group = c("g1", "g1", "g2", "g2", "g1", "g1") 
)

ggplot()  
  geom_point(data=df, aes(x=x, y=y, group=group, color=group))  
  geom_step(data=df, aes(x=x, y=y, group=group, color=group), size=1)  
  scale_y_continuous(breaks = round(seq(0, 100, by = 5), 1))  
  coord_cartesian(ylim = c(0, 100))