Get 7 day periods and week numbers of these from grouped variables

[latsyrhccm]

Was wondering if anyone could help me out with a data tidying task.

I'm attempting to count 7 day periods in my dataset, and mark each as a week number (1-n) for each grouping.

So far my best attempt was repeating a sequence of 1:7 over the dataset, but it does not account for grouping changes. My desired output would be like so:

site    patient_id    date    time    nebs_day    days
01    R01002     2019-04-04       NA        0     1
01    R01002     2019-04-05 16:57:14        1     2
01    R01002     2019-04-06       NA        0     3
01    R01002     2019-04-07 18:19:49        1     4
01    R01002     2019-04-08       NA        0     5
01    R01002     2019-04-09 15:06:44        1     6
01    R01002     2019-04-10 15:36:47        1     7
01    R01002     2019-04-11       NA        0     1
01    R01002     2019-04-12 17:42:32        1     2
01    R01002     2019-04-13 20:24:46        1     3 
01    R01003     2018-09-06 17:58:41        1     1 # Changes group with patient_id
01    R01003     2018-09-07 00:00:05        3     2
01    R01003     2018-09-08 12:52:57        1     3
01    R01003     2018-09-09 00:31:59        4     4
01    R01003     2018-09-10 12:36:52        2     5
01    R01003     2018-09-11 09:39:30        2     6
01    R01003     2018-09-12 09:38:22        3     7

For week numbers, I'm trying to get an output based on days that would look like so:

site    patient_id    date    time    nebs_day    days  week
01    R01002     2019-04-04       NA        0     1    34 #Increases by one for each period
01    R01002     2019-04-05 16:57:14        1     2    34
01    R01002     2019-04-06       NA        0     3    34
01    R01002     2019-04-07 18:19:49        1     4    34
01    R01002     2019-04-08       NA        0     5    34
01    R01002     2019-04-09 15:06:44        1     6    34
01    R01002     2019-04-10 15:36:47        1     7    34 
01    R01002     2019-04-11       NA        0     1    35 
01    R01002     2019-04-12 17:42:32        1     2    35
01    R01002     2019-04-13 20:24:46        1     3    35 #Not a full 7days, but mark as 1 week
01    R01003     2018-09-06 17:58:41        1     1    1  #Count resets with new patient_id
01    R01003     2018-09-07 00:00:05        3     2    1
01    R01003     2018-09-08 12:52:57        1     3    1
01    R01003     2018-09-09 00:31:59        4     4    1
01    R01003     2018-09-10 12:36:52        2     5    1
01    R01003     2018-09-11 09:39:30        2     6    1
01    R01003     2018-09-12 09:38:22        3     7    1

Any help with this would be greatly appreciated. I have tried using days() and weeks() prior to this, but it's a requirement to distinguish 7 day periods in my dataset so it would not be fully accurate using lubridate.

Here's some sample data:

sample <- data.frame(
site = rep(1, each = 17), 
patient_id = c("R01002", "R01002", "R01002", "R01002", "R01002", "R01002", 
"R01002", "R01002", "R01002", "R01002", "R01003", "R01003", "R01003",
"R01003", "R01003", "R01003", "R01003"), 
date = c("2019-04-04", "2019-04-05","2019-04-06", "2019-04-07", "2019-04-08", "2019-04-09", "2019-04-10","2019-04-11", "2019-04-12", "2019-04-13", 
"2018-09-06", "2018-09-07", "2018-09-08", "2018-09-09", "2018-09-10", 
"2018-09-11", "2018-09-12"), 
nebs_day = c(0,1,0,1,0,1,1,0,1,1,1,3,1,4,2,2,3))

Thanks!

[williaml]

There is a week function in lubridate, but not sure how it would work with your updated count.

e.g.

library(tidyverse)
library(lubridate)

sample %>% 
  mutate(week = week(date))