ggplot2 multiline spline smoothing-CodePudding

How do I do spline-smoothing on a multi-line plot in the code segment below? The attached figure shows the two plots generated by this code.

Thanks!

library(tidyverse)
library(ggplot2)

# group by cyl
by_cyl <- mtcars %>% group_by(cyl) %>% summarize(mpg = mean(mpg))

# group by cyl & gear
by_cyl_gear <- mtcars %>% group_by(cyl, gear) %>% summarize(mpg = mean(mpg))

# spline smoothing for a single line plot
spline_d <- as.data.frame(spline(by_cyl$cyl, by_cyl$mpg))
ggplot(data = by_cyl, aes(x = cyl, y = mpg))  
    geom_line(data = spline_d, aes(x = x, y = y))  
    geom_point()  
    ggtitle("Unpartitioned (smooth)")

# multi-line plot - how to do spline-smoothing here?
by_cyl_gear$gear <- as.character(by_cyl_gear$gear)
ggplot(data = by_cyl_gear, aes(x = cyl, y = mpg))  
    geom_line(aes(color = gear))  
    geom_point(aes(color = gear))  
    ggtitle("Partitioned by gear (non-smooth)")  
    theme(legend.position = c(0.9, 0.6))

CodePudding user response：

Basically it's the same as for a single line. You could split by gear, use lapply to compute the spline for each split and then bind back together using e.g. bind_rows:

library(dplyr)
library(ggplot2)

# group by cyl & gear
by_cyl_gear <- mtcars %>% 
  group_by(cyl, gear) %>% 
  summarize(mpg = mean(mpg)) %>% 
  mutate(gear = as.character(gear))
#> `summarise()` has grouped output by 'cyl'. You can override using the `.groups` argument.

# split, lapply, bind_rows
spline_d <- by_cyl_gear %>% 
  split(.$gear) %>% 
  lapply(function(x) as.data.frame(spline(x$cyl, x$mpg))) %>% 
  bind_rows(.id = "gear")

ggplot(data = by_cyl_gear, aes(x = cyl, y = mpg, color = gear))  
  geom_line(data = spline_d, aes(x = x, y = y))  
  geom_point()  
  ggtitle("Partitioned by gear (non-smooth)")  
  theme(legend.position = c(0.9, 0.6))