订购周数是移动时间段的因素

Question

I have a nice little graph that tracks data over time. The data covers the past 12 weeks rolling, is collected daily, grouped by week and output weekly. It's worked flawlessly until the change in year sent my factor(Week) out the window and started plotting 2019 week 1 ahead of 2018 week 43. How can I automate the factor levels as the order changes crossing the new year? I can manually adjust the factor levels until week 12,but that seems like a bandaid. This is the output I'm trying to achieve without the manual intervention.

data <-  structure(list(ID = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
    1L, 1L, 1L, 1L, 2L, 2L, 2L, 2L, 2L, 3L, 3L, 3L, 3L, 3L, 3L, 3L, 
    3L, 3L, 4L, 4L, 4L, 4L, 4L, 4L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 5L, 
    5L, 5L, 5L, 5L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 6L, 
    7L, 7L, 7L, 7L, 7L, 7L, 7L, 7L, 7L, 7L, 7L, 7L, 8L, 8L, 8L, 8L, 
    8L), .Label = c("58", "66", "68", "77", "79", "80", "84", "98"
    ), class = "factor"), Year = c(2019, 2019, 2019, 2018, 2018, 
    2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 
    2018, 2019, 2019, 2019, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 
    2018, 2018, 2018, 2018, 2018, 2019, 2019, 2019, 2018, 2018, 2018, 
    2018, 2018, 2018, 2018, 2018, 2018, 2019, 2019, 2019, 2018, 2018, 
    2018, 2018, 2018, 2018, 2018, 2018, 2018, 2019, 2019, 2019, 2018, 
    2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 2018, 
    2018, 2018), Week = c(1L, 2L, 3L, 43L, 44L, 45L, 46L, 47L, 48L, 
    49L, 50L, 51L, 43L, 44L, 45L, 46L, 47L, 1L, 2L, 3L, 46L, 47L, 
    48L, 49L, 50L, 51L, 44L, 45L, 46L, 47L, 48L, 49L, 1L, 2L, 3L, 
    43L, 44L, 45L, 46L, 47L, 48L, 49L, 50L, 51L, 1L, 2L, 3L, 43L, 
    44L, 45L, 46L, 47L, 48L, 49L, 50L, 51L, 1L, 2L, 3L, 43L, 44L, 
    45L, 46L, 47L, 48L, 49L, 50L, 51L, 43L, 44L, 45L, 46L, 47L), 
        Output = c(49.9, 67.42, 31.27, 65.8925, 79.2925, 103.47, 
        120.1125, 122.645, 109.3925, 91.3125, 81.06, 78.38, 119.13, 
        93.71, 149.74, 122.8775, 117.1075, 32.81, 29.49, 16.71, 42.8725, 
        60.6425, 71.2, 86.155, 78.6225, 81.605, 73.51, 84.42, 105.4, 
        74.515, 57.825, 42.8475, 11.26, 44.34, 22.83, 41.695, 75.77, 
        80.785, 118.175, 131.2875, 124.2375, 124.905, 85.7275, 61.7525, 
        188.23, 108.42, 42.42, 111.41, 79.8825, 70.4075, 72.695, 
        61.235, 58.7825, 47.9275, 46.5275, 48.1775, 11.09, 10.13, 
        0, 21.3375, 30.1275, 32.97, 53.51, 51.09, 40.385, 35.9025, 
        23.44, 21.1125, 111.94, 102.76, 105.71, 112.36, 115.7875)), class = "data.frame", row.names = c(NA, 
    -73L))

 ggplot(data, 
       aes(factor(Week,
                  levels = c(43, 44, 45, 46, 47, 48, 49, 50, 51, 1, 2, 3)),
           Output)) +
geom_line(size = 1.5, 
          aes(colour = ID,
              group = ID)) +
geom_point(aes(y = Output,
               colour = ID),
           size = 4,
           shape = 21,
           fill = "black",
           stroke = 2) + 
  labs(x = "Week Number",
       y = "output") +
  scale_y_continuous(breaks = pretty_breaks(n = 20),
                     limits = c(0, NA))

Answer 1

For base R, consider creating a date column which involves formatting year and then adding corresponding week to converted date. From there, plot with scale_x_date() formatted to week number:

data$Date <- with(data, as.Date(paste0(Year, "-01-01"), format="%Y-%m-%d", origin="1970-01-01") +
                               as.difftime(Week, unit="weeks")
             )

head(data)
#   ID Year Week   Output       Date
# 1 58 2019    1  49.9000 2019-01-08
# 2 58 2019    2  67.4200 2019-01-15
# 3 58 2019    3  31.2700 2019-01-22
# 4 58 2018   43  65.8925 2018-10-29
# 5 58 2018   44  79.2925 2018-11-05
# 6 58 2018   45 103.4700 2018-11-12

ggplot(data, 
       aes(Date,
           Output)) +
  geom_line(size = 1.5, 
            aes(colour = ID,
                group = ID)) +
  geom_point(aes(y = Output,
                 colour = ID),
             size = 4,
             shape = 21,
             fill = "black",
             stroke = 2) + 
  labs(x = "Week Number",
       y = "output") +
  scale_y_continuous(breaks = pretty_breaks(n = 20),
                     limits = c(0, NA)) +
  scale_x_date(breaks = pretty_breaks(n = 20),
                   labels = date_format("%W"))

Answer 2

Per my comment, I changed the combination of week/year into a date, created the plot, and then formatted the x-axis back as a week:

library(tidyverse)
library(lubridate)

data %>% 
  mutate(date = dmy(paste0("0101", Year)) + days(7*(Week)),
         date = floor_date(date, "weeks", week_start = 1)) %>%   # spoof week/year into date
  ggplot(aes(date, Output)) +
  geom_line(size = 1.5, 
            aes(colour = ID,
                group = ID)) +
  geom_point(aes(y = Output,
                 colour = ID),
             size = 4,
             shape = 21,
             fill = "black",
             stroke = 2) + 
  labs(x = "Week Number",
       y = "output") %>% 
  scale_x_date(date_breaks = "weeks", date_labels = "%W")

Hopefully you have the original dates rather than having to shoehorn the week number back into a date, since this is a little messy.