R - 根据连续行和因子级别构建唯一组

Question

In general, how would I group based on identical factors as long as they come from consecutive rows in the data frame? For example, I would like the desired good_output below from test .

test <- data.frame(time = 1:10, letter = c("a","a","a","b","a","a","a","b","b","b"))
bad_output <- test %>% group_by(letter) %>% summarize(mean_time = mean(time))
bad_output
# A tibble: 2 x 2
  letter mean_time
  <fct>      <dbl>
1 a           4   
2 b           7.75

good_output <- data.frame(letter=c("a","b","a","b"), id=c(1,1,2,2), mean_time=c(2,4,6,9))
good_output
  letter id mean_time
1      a  1         2
2      b  1         4
3      a  2         6
4      b  2         9

Answer 1

We can do a group by on 'letter' and the run-length-id ( rleid from data.table ) on the 'letter', summarise to get the mean of 'time', create the sequence column with row_number() and select out the 'grp' column

library(dplyr)
library(data.table)
test %>% 
    group_by(letter, grp = rleid(letter))  %>%
    summarise(mean_time = mean(time)) %>%       
    mutate(id = row_number()) %>%
    ungroup %>%
    select(-grp)
# A tibble: 4 x 3
#  letter mean_time    id
#  <fct>      <dbl> <int>
#1 a              2     1
#2 a              6     2
#3 b              4     1
#4 b              9     2