Especially since I had to start working as a freelancer, it became essential to keep an accurate record of how much time I spend on each task and project. I checked out some software tools available online (namely TimeCamp, which I’ve used for a while), but I decided to see if I could write my own R functions for tracking time use, especially when working offline (some of us still like or need to work offline sometimes!), and for producing some related plots and analyses afterwards. Here’s what I came up with.

I first need to have a file named “PROJECTS.csv“, with a column named
“project” and a column named “task“, and the names of the projects and tasks I’ll be working on, like this:

I can manually add projects and tasks to this table at any time. Then I use the ‘start_task‘ and ‘stop_task‘ functions below to add entries to another file, which is named “DONT-TOUCH_timeman_data.csv“, has five columns (proj, task, start, stop, time) and should not be moved or modified manually:

projects <- read.csv("PROJECTS.csv")
tail(projects)

timeman_data <- read.csv("DONT-TOUCH_timeman_data.csv", as.is = TRUE)
tail(timeman_data)

start_task <- function(proj, task, time = NULL) {
  if (nrow(timeman_data) > 0 && is.na(timeman_data[nrow(timeman_data), ncol(timeman_data)]))
    stop ("previous task must be stopped before starting another")
  if (!(proj %in% unique(projects$project)))
    stop ("proj must exist in 'PROJECTS.csv'")
  if (!(task %in% unique(projects$task)))
    stop ("task must exist in 'PROJECTS.csv'")
  #if (!(proj %in% unique(timeman_data$projecto)))
    #timeman_data[nrow(timeman_data) + 1, "projecto"] <- proj
  if (is.null(time))
    time <- Sys.time()
  timeman_data[nrow(timeman_data) + 1, 1:3] <- c(proj, task, time)
  timeman_data <<- timeman_data
}

stop_task <- function(time = NULL) {
  if (nrow(timeman_data) > 0 && !is.na(timeman_data[nrow(timeman_data), ncol(timeman_data)]))
    stop ("task must be started before being stopped")
  if (is.null(time))
    time <- Sys.time()
  #this_row <- which.max(timeman_data$projecto == proj)
  this_row <- nrow(timeman_data)
  timeman_data[this_row, "stop"] <- as.character(time)
  timeman_data[this_row, "time"] <- difftime(timeman_data[this_row, "stop"], timeman_data[this_row, "start"])
  timeman_data <<- timeman_data
  write.csv(timeman_data, "DONT-TOUCH_timeman_data.csv", row.names = FALSE)
  cat("Showing last rows in tasks table:\n\n")
  tail(timeman_data)
}

current_task <- function() {
  timeman_data[nrow(timeman_data), ]
}

I need to use ‘start_task‘ every time I start working on a task, and then ‘stop_task‘ every time I stop. This adds an entry with the starting, stopping, and total time spent in each instance:

unique(projects$proj)
unique(projects$task)
unique(timeman_data$proj)

# ADD START:
start_task(proj = "timeman", task = "debugging")
tail(timeman_data)

# ADD STOP:
stop_task()
tail(timeman_data)

I can then apply all sorts of R analyses and plots to the time-tracking data, which is recorded in the timeman_data data frame and CSV file. However, these functions are certainly far from ideal, as they use “unsafe” operators (like ‘<<-‘) and they rely on a file remaining untouched in its folder. It’s plain to see how things could easily go wrong at any moment…. Suggestions welcome on how this could be improved and made safer! While this was an interesting exercise, I think I’ll just stick to TimeCamp for the time being