data transformation using R [closed] - r

Question

Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 1 year ago.
Improve this question
I got some data such as this
structure(list(id = c(1, 1, 1, 1, 2, 2, 3, 3, 3, 3, 3, 3), dead = c(1,
1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0), futime = c(2062, 2062, 2062,
2062, 2151, 2151, 388, 388, 388, 388, 388, 388), event = c("hosp",
"out", "hosp", "out", "hosp", "out", "hosp", "out", "hosp", "out",
"hosp", "out"), event_time = c(36, 52, 775, 776, 1268, 1283,
178, 192, 271, 272, 387, 377.9)), class = "data.frame", row.names = c(NA,
-12L))
and I would like to make it look like this
structure(list(id2 = c(1, 1, 1, 1, 1, 2, 2, 2, 3, 3, 3, 3, 3,
3, 3), dead2 = c(NA, NA, NA, NA, 1, NA, NA, 1, NA, NA, NA, NA,
NA, NA, NA), futime2 = c(NA, NA, NA, NA, 2062, NA, NA, 2151,
NA, NA, NA, NA, NA, NA, 388), event2 = c("hosp", "out", "hosp",
"out", "death", "hosp", "out", "death", "hosp", "out", "hosp",
"out", "hosp", "out", "censored"), event_time2 = c(36, 52, 775,
776, 2062, 1268, 1283, 2151, 178, 192, 271, 272, 387, 377.9,
388)), class = "data.frame", row.names = c(NA, -15L))
So basically, I want the dead == 1 and the value in the futime column to appear in the last observation for each id. and create a new column where all events are sequentially entered. Thanks

Answer 1

I didn't put the "2"s in the column names of the result, but you can make that change easily enough if you need it.
library(dplyr)
last_rows = df %>%
select(id, dead, futime) %>%
group_by(id) %>%
slice(1) %>%
mutate(
event = ifelse(dead == 1, "death", "censored"),
event_time = futime
)
result = df %>%
mutate(
dead = NA,
futime = NA
) %>%
bind_rows(last_rows) %>%
arrange(id, event_time)
result
# id dead futime event event_time
# 1 1 NA NA hosp 36.0
# 2 1 NA NA out 52.0
# 3 1 NA NA hosp 775.0
# 4 1 NA NA out 776.0
# 5 1 1 2062 death 2062.0
# 6 2 NA NA hosp 1268.0
# 7 2 NA NA out 1283.0
# 8 2 1 2151 death 2151.0
# 9 3 NA NA hosp 178.0
# 10 3 NA NA out 192.0
# 11 3 NA NA hosp 271.0
# 12 3 NA NA out 272.0
# 13 3 NA NA out 377.9
# 14 3 NA NA hosp 387.0
# 15 3 0 388 censored 388.0

Answer 2

Here is one method using group_modify and add_row
library(dplyr)
library(tibble)
df1 %>%
group_by(id, futime) %>%
group_modify(~ .x %>%
add_row(dead = NA^!last(.x$dead), event_time = last(.y$futime),
event = if(last(.x$dead) == 1) "death" else "censored")) %>%
mutate(across(c(dead), ~ replace(., row_number() != n(), NA))) %>%
group_by(id) %>%
mutate(futime = replace(futime, duplicated(futime, fromLast = TRUE),
NA)) %>%
ungroup
-output
# A tibble: 15 × 5
id futime dead event event_time
<dbl> <dbl> <dbl> <chr> <dbl>
1 1 NA NA hosp 36
2 1 NA NA out 52
3 1 NA NA hosp 775
4 1 NA NA out 776
5 1 2062 1 death 2062
6 2 NA NA hosp 1268
7 2 NA NA out 1283
8 2 2151 1 death 2151
9 3 NA NA hosp 178
10 3 NA NA out 192
11 3 NA NA hosp 271
12 3 NA NA out 272
13 3 NA NA hosp 387
14 3 NA NA out 378.
15 3 388 NA censored 388