i have data like so:
| Date of Dose 2 | episode_1 | episode_2 | episode_3 | episode_4 | episode_5 |
|---|---|---|---|---|---|
| 2022-01-15 | 2022-02-01 | 2021-12-25 | NA | 2022-03-01 | NA |
I need to identify the dates which are 14 days greater than the date of dose 2 from the 5 episode columns. output should be:
| Date of BI |
|---|
| 2022-02-01, 2022-03-01 |
what I have tried so far:
vaccine_received_test <- vaccine_received %>%
rowwise() %>%
mutate(
dose_2_plus_14 = `Date of Dose 2` + days(14),
# Check if any episode date is greater than 'Date of Dose 2 + 14 days'
date_of_BI = case_when(
any(c(episode_1, episode_2, episode_3, episode_4, episode_5) > dose_2_plus_14,
na.rm = TRUE) ~ any(c(episode_1, episode_2, episode_3, episode_4, episode_5),
TRUE ~ NA
)
)
any ideas on how I can achieve this?
3
You could pivot, then filter, then pivot:
library(tidyr)
library(dplyr)
pivot_longer(df, -1) |>
filter(value-`Date of Dose 2`>14) |>
pivot_wider(id_cols=1)
___
# A tibble: 1 × 3
`Date of Dose 2` episode_1 episode_4
<date> <date> <date>
1 2022-01-15 2022-02-01 2022-03-01
Data:
df <- structure(list(`Date of Dose 2` = structure(19007, class = "Date"),
episode_1 = structure(19024, class = "Date"), episode_2 = structure(18986, class = "Date"),
episode_3 = NA, episode_4 = structure(19052, class = "Date"),
episode_5 = NA), row.names = c(NA, -1L), class = c("tbl_df",
"tbl", "data.frame"))
1
With dplyr, you can use select() with where():
library(dplyr)
df %>%
select(where(~ difftime(.x, df[["Date of Dose 2"]]) > 14 & !is.na(.x)))
# # A tibble: 1 × 2
# episode_1 episode_4
# <date> <date>
# 1 2022-02-01 2022-03-01
And a fairly obscure Base
df[, c(1, which(difftime(apply(df[1, ], 1, as.POSIXct)[2:6],
apply(df[1, ], 1, as.POSIXct)[1]) > 14)+1)]
Date of Dose 2 episode_1 episode_4
1 2022-01-15 2022-02-01 2022-03-01
where apply here both provides the above @limey suggested ‘go long’ and transform to POSIXct.