Issue/#169 incorrect count anon users

[WahiduzzamanKhan]

Closes #169

Changes description

• Added prefix to the auto generated username for anonymous users.
• Remove the Total Anonymous Users valueBox from the Analytics app.

Definition of done

• Have you read the Contributing Guidelines?
• NEWS.md file has been updated
• Development version has been bumped (x.y.z.90XX)
• Issue has been linked with this PR (via Closing keywords or right sidebar under Development)

[averissimo]

All anonymous users are being tracked as a single one if track_anonymous = FALSE

diff --git a/inst/examples/app/instrumentation/app.R b/inst/examples/app/instrumentation/app.R
index a6c9722..e7cf8bf 100644
--- a/inst/examples/app/instrumentation/app.R
+++ b/inst/examples/app/instrumentation/app.R
@@ -112,7 +112,8 @@ counter_server <- function(id) {
 shinyApp(ui = ui, server = function(input, output, session) {
   telemetry$start_session(
     track_values = TRUE,
-    navigation_input_id = "uisidebar"
+    navigation_input_id = "uisidebar",
+    track_anonymous_user = FALSE
   )
 
   # server code

I think this could be solved by either:

1. When counting users: add every unique session without user (is.na(.data$username))
   • see below
2. On Telemetry$get_user the default value returns a unique string (such as paste0("anon_user_", session$token) or a generated one)

total_users <- length(
  unique(
    selected_log_data() %>%
      dplyr::filter(.data$type == "login") %>%
      dplyr::filter(.data$username != "") %>%
      dplyr::pull(.data$username)
  )
) + 
  length(
    unique(
      selected_log_data() %>%
        dplyr::filter(.data$type == "login") %>%
        dplyr::filter(is.na(.data$username)) %>%
        dplyr::pull(.data$session)
    )
  )

[WahiduzzamanKhan]

All anonymous users are being tracked as a single one if track_anonymous = FALSE

diff --git a/inst/examples/app/instrumentation/app.R b/inst/examples/app/instrumentation/app.R
index a6c9722..e7cf8bf 100644
--- a/inst/examples/app/instrumentation/app.R
+++ b/inst/examples/app/instrumentation/app.R
@@ -112,7 +112,8 @@ counter_server <- function(id) {
 shinyApp(ui = ui, server = function(input, output, session) {
   telemetry$start_session(
     track_values = TRUE,
-    navigation_input_id = "uisidebar"
+    navigation_input_id = "uisidebar",
+    track_anonymous_user = FALSE
   )
 
   # server code

I think this could be solved by either:

1. When counting users: add every unique session without user (is.na(.data$username))

   • see below

2. On Telemetry$get_user the default value returns a unique string (such as paste0("anon_user_", session$token) or a generated one)

total_users <- length(
  unique(
    selected_log_data() %>%
      dplyr::filter(.data$type == "login") %>%
      dplyr::filter(.data$username != "") %>%
      dplyr::pull(.data$username)
  )
) + 
  length(
    unique(
      selected_log_data() %>%
        dplyr::filter(.data$type == "login") %>%
        dplyr::filter(is.na(.data$username)) %>%
        dplyr::pull(.data$session)
    )
  )

all the suggested changes have been implemented.

• tracking for anonymous users have been disabled in the instrumentation app
• the username of the anonymous users when track_anonymous_user = FALSE follows the form anon_user_<session$token>

[averissimo]

The diff was just to show how to test the application.

I've reverted it.

All good! Have a great weekend