Flag Completion Status at Event, Session, and/or Applicant Level

Creates binary completion flag columns on a web analytics dataframe, indicating whether an applicant completed the target event (ApplicantSharedIncomeSummary) at the event, session, and/or applicant level.

Usage

flag_completed(df, type = c("event", "session", "applicant"))

Arguments

df

A dataframe containing web analytics data. Must include the following columns:

event: Character. The name of the tracked event.
pilot: Character or factor. The pilot group identifier, used as a grouping variable for session- and applicant-level flags.
cbv_flow_id: Character or numeric. A unique session identifier. Required when "session" is included in type.
distinct_id: Character or numeric. A unique applicant identifier. Required when "applicant" is included in type.

type

Character vector specifying which completion flags to generate. One or more of:

"event": Adds completed_event: TRUE if the row's event is ApplicantSharedIncomeSummary, FALSE otherwise.
"session": Adds completed_session: TRUE if any event within the same pilot + cbv_flow_id session is a completion event.
"applicant": Adds completed_applicant: TRUE if any event across all sessions for the same pilot + distinct_id applicant is a completion event.

Defaults to c("event", "session", "applicant") (all three).

Value

A dataframe with the same rows as df, with one or more of the following columns appended depending on type:

completed_event: Logical. Row-level completion flag. Included only if "event" is in type.
completed_session: Logical. Session-level completion flag. Included only if "session" is in type.
completed_applicant: Logical. Applicant-level completion flag. Included only if "applicant" is in type.

Details

The target completion event is defined internally as "ApplicantSharedIncomeSummary". A row-level completed_event flag is first created, then session- and applicant-level flags are derived by applying any() within each group. any() returns TRUE if at least one value in the group is TRUE, making the aggregation logic explicit and self-documenting.

If "event" is not included in type, the intermediate completed_event column is dropped from the returned dataframe.

Grouping configurations for "session" and "applicant" are defined internally and applied functionally via purrr::reduce(), making it straightforward to extend the function with additional grouping levels in the future.

Examples

if (FALSE) { # \dontrun{
# Add all three flags (default)
df_flagged <- flag_completed(df)

# Add only the applicant-level flag
df_flagged <- flag_completed(df, type = "applicant")

# Add session and applicant flags, but not the raw event-level flag
df_flagged <- flag_completed(df, type = c("session", "applicant"))
} # }