For multiple individuals (ident column), i have multiple instances that are calculated at different time points.
In my real database i have like more than 10 time points and more than 10 instances.
Here as the example i only use 2 time points (M0 & M3) and 2 instances (abc & def):
I would like to keep the ident column and pivot_longer to have 2 columns after : abc and def . I succeed to do it if it’s only for one instance, with the names_pattern options, but i struggle to apply that to more than one instance. I tried by having values_to=c("abc",def") but then i don’t know what to put in the regex pattern.
data = data.frame(ident=c("A","B","C","D","E"),
M0_abc=c(1,0,1,0,1),
M0_def=c(1,1,1,1,0),
M3_abc=c(1,1,0,0,0),
M3_def=c(0,1,0,1,0))
> data
ident M0_abc M0_def M3_abc M3_def
1 A 1 1 1 0
2 B 0 1 1 1
3 C 1 1 0 0
4 D 0 1 0 1
5 E 1 0 0 0
data_long <- data %>% select(ident, contains("abc")) %>%
pivot_longer(cols=contains("abc"),
names_to="time",
values_to="abc",
names_pattern = "(.*)_abc")
> data_long
# A tibble: 10 × 3
ident time abc
<chr> <chr> <dbl>
1 A M0 1
2 A M3 1
3 B M0 0
4 B M3 1
5 C M0 1
6 C M3 0
7 D M0 0
8 D M3 0
9 E M0 1
10 E M3 0
>Solution :
In case of multiple value columns you could use the special ".value" like so:
library(tidyr)
library(dplyr, warn = FALSE)
data %>%
pivot_longer(
-ident,
names_to = c("time", ".value"),
names_pattern = "(.*)_(.*)"
)
#> # A tibble: 10 × 4
#> ident time abc def
#> <chr> <chr> <dbl> <dbl>
#> 1 A M0 1 1
#> 2 A M3 1 0
#> 3 B M0 0 1
#> 4 B M3 1 1
#> 5 C M0 1 1
#> 6 C M3 0 0
#> 7 D M0 0 1
#> 8 D M3 0 1
#> 9 E M0 1 0
#> 10 E M3 0 0