Continuity Creator

continuity is based on the 'islands & gaps' concept. Use case for continuity is to roll up a sequence consisting of smaller duration into a larger epoch governed by some meaningful separation between \(n_\text{lower}\) and \({n - 1}_\text{upper}\) segments.

continuity(
  data,
  map_fields,
  time_fields,
  timeout = 0,
  boundary_name = "window",
  archipelago = TRUE,
  show.all = FALSE
)

Arguments

data

(object): The source dataset, including all non-sessioning fields desired

map_fields

(string): A vector of strings or symbols indicating the field names that will partition data

time_fields

(string): A vector of strings or symbols indicating the field names to use as "start" and "stop" temporal indices. If only one value is given, that value will be repeated as the "stop" index

timeout

The largest allowable 'gap' in a series of time values before a new 'island' begins: can be a quoted expression that conditionally determines the value. If using date or datetime values for time_fields, specify the timeout using an appropriate lubridate functions (e.g., days).

boundary_name

(string): The name root of the boundary column names (e.g., "episode" => "episode_start_idx", "episode_end_idx")

archipelago

(logical | TRUE): Should the output include the islands and gaps generated?

show.all

(logical | FALSE): Should the output include all of the columns of the output? show.all and archipelago are independent

Value

A data.table with the following columns <mapFields>, <X>, <Y>:

X (archipelago = TRUE)

{boundary_name}_start_idx, {boundary_name}_end_idx, ISLAND, GAP

Y (show.all = TRUE)

island_idx, seq_idx, start_idx, stop_idx, rec_idx, map_partition, delta_start, delta_stop