Compute daily entry and exit flows

Calculate daily counts of entity registrations (entries) and deletions (exits) classified by industry (NACE code) and geography (municipality code). Three data paths, selected automatically:

Usage

brreg_flows(
  data = NULL,
  updates = NULL,
  by = c("nace_1", "municipality_code"),
  from = NULL,
  to = NULL,
  legal_form = NULL
)

Arguments

data

Optional. A tibble from brreg_download() or a snapshot. Required when no changelog exists. Must contain org_nr, registration_date, nace_1, and municipality_code.

updates

Optional. A tibble from brreg_updates() with CDC events.

by

Character vector of grouping columns. Default c("nace_1", "municipality_code"). Use NULL for national totals, or any column present in data.

from, to

Date range for the output. NULL defaults to the range of observed events.

legal_form

Optional character vector of legal form codes to include (e.g. c("AS", "ENK")). NULL includes all.

Value

A tibble with columns: date (Date), grouping columns from by, entries (integer), exits (integer), net (integer: entries - exits). An attribute flow_source records which data sources contributed.

Details

1. Changelog path (preferred) — when brreg_sync() has been run, reads directly from the persistent changelog. Provides timestamped entries, exits, and field-level transitions. No arguments needed.

2. Bulk + CDC path — pass data (from brreg_download()) and optionally updates (from brreg_updates()). Registration dates provide historical entries; CDC provides recent entries + exits.

3. Bulk-only path — pass data alone. Only entries are computed (no exit data available).

Entry vs. founding date

This function uses registration_date (registreringsdatoEnhetsregisteret), NOT founding_date (stiftelsesdato). Registration date is when the entity entered the registry. Founding date can precede registration by months (AS companies) or years (associations).

See also

brreg_download() to get bulk data, brreg_updates() to get CDC events, brreg_series() for snapshot-based time series, as_brreg_tsibble() for tsibble conversion.

Other tidybrreg panel functions: as_brreg_tsibble(), brreg_change_summary(), brreg_changes(), brreg_events(), brreg_panel(), brreg_replay(), brreg_series()

Examples

if (FALSE) { # interactive() && curl::has_internet()
# \donttest{
entities <- brreg_download()
flows <- brreg_flows(entities)

# With CDC exits
cdc <- brreg_updates(since = "2026-01-01", size = 10000)
flows <- brreg_flows(entities, updates = cdc)

# Monthly by NACE section
flows |>
  dplyr::mutate(month = format(date, "%Y-%m"),
                nace_section = substr(nace_1, 1, 2)) |>
  dplyr::summarise(entries = sum(entries), exits = sum(exits),
                   .by = c(month, nace_section))
# }
}