I have a dataframe with each observation representation an object in the real world I have counted using OpenCV. After some mutations and using tidyverse and lubridate my dataframe looks like this:

> head(odcCountsRoadUsers)
       frameId objectClass dayOfWeek ymdhms              hourOfDay
  1    1133    car         1         2021-12-05 13:45:25 13
  2    1159    car         1         2021-12-05 13:45:26 13
  3    1243    car         1         2021-12-05 13:45:31 13
  4    1280    person      1         2021-12-05 13:45:33 13
  5    1305    bus         1         2021-12-05 13:45:34 13
  6    1812    person      1         2021-12-05 13:46:03 13

I have data for multiple days, also containing multiple counts for the same dayOfWeek. What I want to do is create a histogram for each day of the week (7 in total) with the number of observations for each hour of the day. The result should look similar to this (one histogram per day of the week):

My problems are:

1. How do I use ggplot2 so that it uses the POSIXct ymdhms as x-axis and for the y-axis does an aggregation of the number of observations by hourOfDay (or directly using ymdhms if that's possible) and filters by dayOfWeek?
2. Since I have multiple days with records I only want the average for a dayOfWeek and not simple add the count of all observations for different dates together. How can I do this efficiently? Does it make sense to create a separate dataframe for that before plotting or can this be done easily using ggplot2?

Bonus question regarding a chance to separate/group by objectClass, but those two are the most pressing questions I cannot seem to figure out on my own.

The

Adding the object class as an argument of group_by allows us for additional stratification:

aggregated_data2 <-
  data %>%
  mutate(
    weekday = wday(time, label = TRUE),
    hour = hour(time),
    date = date(time)
  ) %>%
  count(objectClass, weekday, date, hour) %>%
  # average e.g over all mondays for each object class
  group_by(objectClass, weekday, hour) %>%
  summarise(n = mean(n))

aggregated_data2 %>%
  ggplot(aes(hour, n, color = objectClass))  
  geom_line()  
  facet_wrap(~weekday, scales = "free")