How to draw regression line R ggplot-CodePudding

I have the following dataframe

df_acc_score = data.frame(group  = c("con", "dem", "sch"),
                          cohort = c("a1", "a3"), 
                          score = c(12, 15, 10, 16, 19, 15, 18, 20, 17, 20, 24, 19))

I would like to draw the line between the accuracy mean score in a1 and a3 for each group. For now, my code for the graph is the following

ggplot(df_acc_score, aes(x = cohort, y = score, color = group)) 
  geom_hline(yintercept = 5, color = "grey") 
  stat_summary(fun = "mean", 
               size = 0.1) 
  theme_bw() 
  theme(strip.background = element_blank()) 
  ylab("Mean accuracy score") 
  xlab("Cohort")

How shall I do to add the line between the mean score in a1 and mean score in a3 for each group?

Thanks a lot for your response

CodePudding user response：

First, your data is not complete which means your group and cohort variables have fewer values than the score. Also a regression line with only two points is not ideal. A different option to show mean points is by calculating them beforehand so you can show them with geom_point and combine them with geom_line. Here is a reproducible example:

library(dplyr)
library(ggplot2)
df_acc_score %>%
  group_by(group, cohort) %>%
  mutate(mean_cohort = mean(score)) %>%
  ggplot(aes(x = cohort, y = score, color = group, group = group))  
  geom_hline(yintercept = 5, color = "grey") 
  geom_line(size = 0.1)  
  geom_point()  
  theme_bw() 
  theme(strip.background = element_blank()) 
  ylab("Mean accuracy score") 
  xlab("Cohort")