Same column ( different row ) operations in R - r

Question

I have a big database and I'm trying to create a new column starting from an existing one doing the difference between elements in consecutive cells ( same column, different row):
existing_column
new_column
A
A-B
B
B-C
C
C-D
D
D-E
...
...
Z
Z-NULL
The way I'm doing it is to duplicate existing column into a dummy one, remove first element, adding NULL as last element and subtracting the existing column and the dummy one ... is there a better way? Thank you
exist <-c("A","B","C","D","E")
db<-data.frame(exist)
dummy<-exist[-1]
dummy[length(dummy)+1]<-"NULL"
new_col<-paste(exist,"-",dummy)
new_col
db<-data.frame(exist,new_col)
db

Answer 1

Does this work:
library(dplyr)
df <- data.frame(existing_column = LETTERS)
df %>% mutate(new_column = paste(existing_column, lead(existing_column, default = 'NULL'), sep = '-'))
existing_column new_column
1 A A-B
2 B B-C
3 C C-D
4 D D-E
5 E E-F
6 F F-G
7 G G-H
8 H H-I
9 I I-J
10 J J-K
11 K K-L
12 L L-M
13 M M-N
14 N N-O
15 O O-P
16 P P-Q
17 Q Q-R
18 R R-S
19 S S-T
20 T T-U
21 U U-V
22 V V-W
23 W W-X
24 X X-Y
25 Y Y-Z
26 Z Z-NULL

Answer 2

Try the code below
transform(
df,
new_column = paste(existing_column, c(existing_column[-1], NA), sep = "-")
)
which gives
existing_column new_column
1 A A-B
2 B B-C
3 C C-D
4 D D-E
5 E E-F
6 F F-G
7 G G-H
8 H H-I
9 I I-J
10 J J-K
11 K K-L
12 L L-M
13 M M-N
14 N N-O
15 O O-P
16 P P-Q
17 Q Q-R
18 R R-S
19 S S-T
20 T T-U
21 U U-V
22 V V-W
23 W W-X
24 X X-Y
25 Y Y-Z
26 Z Z-NA

Answer 3

If you are working with numeric data just represented as characters in your example, you can use mutate() and lead()
df<-data.frame(old_col=sample(1:10))
df%>%mutate(new_col=old_col-lead(old_col, default = 0))
old_col new_col
1 10 4
2 6 -3
3 9 8
4 1 -1
5 2 -5
6 7 3
7 4 1
8 3 -5
9 8 3
10 5 5

Answer 4

In case there is a need of a fast data.table version
dt[, new_column:=paste(exist, shift(exist, type="lead"), sep="-")]
Edit. Turns it isn't much faster:
df = data.table(exist = rep(letters, 80000))
> m = microbenchmark::microbenchmark(
... a = df %>% mutate(new_column = paste(exist, lead(exist, default = 'NULL'), sep = '-')),
...
... b = transform(
... df,
... new_column = paste(exist, c(exist[-1], NA), sep = "-")
... ),
...
... d = df[, new_column := paste(exist, shift(exist, type="lead"), sep="-")]
... )
> m
Unit: milliseconds
expr min lq mean median uq max neval
a 292.2430 309.6150 342.0191 323.9778 361.0937 603.8449 100
b 349.4509 383.3391 475.0177 423.8864 472.0276 2136.2970 100
d 294.6786 302.8530 332.3989 315.6228 340.9642 641.8345 100