Station data

Add the packages that you’ll use:

library(lubridate)
library(tidyverse)

I am creating this test data here, and hiding the code to keep it simple :)

bib	gender	age	cun_start	cun_end	sec_start	sec_end
1	male	34	01:52	02:01	01:52	02:01
2	male	33	01:53	01:55	01:53	01:55
3	female	46	01:51	01:52	01:51	01:52
4	male	46	01:53	01:56	01:53	01:56
5	male	28	01:52	01:56	01:52	01:56
6	female	45	01:51	01:53	01:51	01:53
7	male	38	01:53	01:54	01:53	01:54
8	male	31	01:51	01:53	01:51	01:53
9	female	26	01:51	01:53	01:51	01:53
10	female	42	01:53	01:54	01:53	–|–

To process your stations data:

• Convert all of the start and end times from characters to time in hours minutes and seconds
• Calculate the time spent at the station, in seconds

See code, followed by explanation below

# create a new dataframe, called stations from data
stations <- data |> 
  mutate_at(vars(-bib, -gender, -age), ~hm(.)) |> 
  mutate(cun_diff = period_to_seconds(cun_end - cun_start)/60,
         sec_diff = period_to_seconds(sec_end - sec_start)/60) 

## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `sec_end = (structure(function (..., .x = ..1, .y = ..2, . =
##   ..1) ...`.
## Caused by warning in `.parse_hms()`:
## ! Some strings failed to parse

Explanation of the above code:

Convert all of the start and end times from characters to time in hours minutes and seconds

• mutate_at: This function is used to mutate specific columns in a data frame.
• vars(-bib, -gender, -age): This part specifies the columns to be mutated, excluding the “bib,” “gender,” and “age” columns. The - sign indicates exclusion.
• ~hm(.): This part is a formula using the tilde (~) to define an anonymous function.
  • hm is a function from the lubridate package that parses hours and minutes.
  • The . inside the formula refers to the current column being operated on.
  • So, ~hm(.) is saying “apply the hm function to each value in the selected columns.” The dot (.) is a placeholder for the column values.

Calculate the time spent at the station, in seconds

• cun_end - cun_start: This calculates the time difference between the “cun_end” and “cun_start” columns, resulting in a period object representing the time duration.
• period_to_seconds(…): This function converts the time duration from the previous step into seconds which fixes the issue when the answer was given in 1H -58M format
• /60: This divides the duration in seconds by 60 to convert it to minutes (OPTIONAL)

Now you have:

kable(stations)

bib	gender	age	cun_start	cun_end	sec_start	sec_end	cun_diff	sec_diff
1	male	34	1H 52M 0S	2H 1M 0S	1H 52M 0S	2H 1M 0S	9	9
2	male	33	1H 53M 0S	1H 55M 0S	1H 53M 0S	1H 55M 0S	2	2
3	female	46	1H 51M 0S	1H 52M 0S	1H 51M 0S	1H 52M 0S	1	1
4	male	46	1H 53M 0S	1H 56M 0S	1H 53M 0S	1H 56M 0S	3	3
5	male	28	1H 52M 0S	1H 56M 0S	1H 52M 0S	1H 56M 0S	4	4
6	female	45	1H 51M 0S	1H 53M 0S	1H 51M 0S	1H 53M 0S	2	2
7	male	38	1H 53M 0S	1H 54M 0S	1H 53M 0S	1H 54M 0S	1	1
8	male	31	1H 51M 0S	1H 53M 0S	1H 51M 0S	1H 53M 0S	2	2
9	female	26	1H 51M 0S	1H 53M 0S	1H 51M 0S	1H 53M 0S	2	2
10	female	42	1H 53M 0S	1H 54M 0S	1H 53M 0S	NA	1	NA

Next steps:

• calculate the mean time difference for each station to impute the data when there is an incorrect –|– value
  • you’ll need to use the rm.na = T parameter so that the NAs don;t mess up your calculation
• replace incorrect –|– with mean
• reprocess the corrected data using the script above
• for each station to calculate:
  • number of runners that left this station
  • number of male runners that left this station
  • number of male runners that left this station
  • average age
  • average age of men
  • average age of women
• Then join this data to the stations spatial dataframe so you can map it