Temporal profile (counts by time bucket)

Usage

entropia_temporal_profile(x, date_var, unit = "month", by = NULL)

Arguments

x

A data frame or tibble of collected rows.

date_var

A column of POSIXct (or Date) timestamps, selected by name or bare (tidyselect).

unit

The time bucket: "second", "minute", "hour", "day", "week" (Monday-start), "month", "quarter" or "year". Default "month".

by

Optional column(s) to break the counts by, selected by name or bare (tidyselect). NULL (default) produces a single time series.

Value

A tibble with the bucket column, the by columns (when given), and a count column n.

Details

Counts the rows of a collected tibble by a date/time column, bucketed to a chosen time unit. This is an R-side analysis helper: pass a materialised tibble (e.g. the output of entropia_collect(), where timestamps are POSIXct). NA timestamps are excluded – a row with an unknown date cannot be placed in time; grouping columns from by keep NA as their own group, following dplyr semantics.

The returned bucket column carries the name of date_var and holds the start of each unit (POSIXct, e.g. midnight for unit = "day"). Rows are ordered by the by columns then the bucket, deterministically. Weeks start on Monday.

Examples

con <- entropia_connect(system.file("extdata", "entropia-example.sqlite",
  package = "entropiaR"
))
items <- entropia_collect(entropia_items(con))
entropia_temporal_profile(items, created_at, unit = "month")
#> # A tibble: 1 × 2
#>   created_at              n
#>   <dttm>              <int>
#> 1 2026-01-01 00:00:00     3
entropia_disconnect(con)