CODEWITHSUNDEEP
r
dzadi
dzadi

Reputation: 137

remove the first row for each group

suppose I have a dataset like this 

df <- data.frame(group = c(rep(1,3),rep(2,2), rep(3,2),rep(4,3),rep(5, 2)), score = c(30, 10, 22, 44, 6, 5, 20, 35, 2, 60, 14,5)) 

   group score
1      1    30
2      1    10
3      1    22
4      2    44
5      2     6
6      3     5
7      3    20
8      4    35
9      4     2
10     4    60
11     5    14
12     5     5

I want to remove the first row for each group, the expected out put should look like this:

   group score
1      1    10
2      1    22
3      2     6
4      3    20
5      4     2
6      4    60
7      5     5

Is there a simple way to do this?

Upvotes: 6

Views: 7613

Answers (5)

MrFlick
MrFlick

Reputation: 206232

Now that dplyr has slice_tail and supports the by= paramter, you can do

df %>% dplyr::slice_tail(n=-1, by=group)

Upvotes: 1

Ronak Shah
Ronak Shah

Reputation: 388982

An option with dplyr is to select rows ignoring 1st row

library(dplyr)
df %>%
  group_by(group) %>%
  slice(2:n())

#  group score
#  <dbl> <dbl>
#1  1.00 10.0 
#2  1.00 22.0 
#3  2.00  6.00
#4  3.00 20.0 
#5  4.00  2.00
#6  4.00 60.0 
#7  5.00  5.00

Another way is shown by @Rich Scriven in now deleted answer 

df %>%
  group_by(group) %>%
  slice(-1)

Upvotes: 16

akira
akira

Reputation: 6117

dplyr::filter(df, group == lag(group))
   group score
1     1    10
2     1    22
3     2     6
4     3    20
5     4     2
6     4    60
7     5     5

See lead and lag of package dplyr for more information:

https://dplyr.tidyverse.org/reference/lead-lag.html

Upvotes: 1

akrun
akrun

Reputation: 887118

Another base R option would be to check the adjacent elements 

df[c(FALSE,df$group[-1]==df$group[-nrow(df)]),]
#   group score
#2      1    10
#3      1    22
#5      2     6
#7      3    20
#9      4     2
#10     4    60
#12     5     5

Here I removed the first observation in 'group' (df$group[-1]) and compared (==) with the vector in which last observation is removed (df$group[-nrow(df)])). As the length of the comparison is one less than the nrow of the dataset, we pad with FALSE at the top and use this as logical index to subset the dataset.

Upvotes: 2

A. Webb
A. Webb

Reputation: 26446

Quite simple with duplicated

df[duplicated(df$group),]

   group score
2      1    10
3      1    22
5      2     6
7      3    20
9      4     2
10     4    60
12     5     5

Upvotes: 9

Related Questions