Prepare data

Overview

To run projections with propop, you need a starting population and projection parameters. You can either use your own data or download them from the Federal Statistical Office (FSO). This vignette explains how to get the data from FSO. You’ll also learn how to prepare the relevant information to run population projections propop.

Note that some data are only available for certain administrative levels.

Required data

If you don’t have the information and data required to run propop::propop() (or propop::project_raw()), you can download most of the data from STAT-TAB. More specifically, the information from the following tables are needed:

Overview of required FSO tables (STAT-TAB)

Table ID	Parameters expressed as…	Variables required for projection
px- x-0104020000_101	number of people (reference scenario)	Inter-cantonal immigration Inter-cantonal emigration International immigration International emigration end of year population size
px- x-0104020000_102	number of people (high growth scenario)	same as 101
px- x-0104020000_103	number of people (low growth scenario)	same as 101
px- x-0104020000_109	rates / probabilities (five scenarios)	Births per mother Mortality International emigration Inter-cantonal emigration Acquisition of Swiss citizenship
px -x-0104020000_106	share of newborns with Swiss nationality born to non-Swiss mothers	Live newborns Live births by age and nationality of the mother (varies between cantons)
Constant parameters not directly available from STAT-TAB are provided as arguments		Start (16) and end (50) of the fertile age of women Proportion of newborns with female sex (100/205)

Convenient way to get FSO data

propop provides two convenience functions to download data from the FSO. These are strongly based on the BFS package and its documentation.

To get the starting population for a spatial unit, you must use the spelling defined in the corresponding FSO table. The entries in the FSO tables may contain special characters. The spelling may also vary between FSO tables.

BFS::bfs_get_metadata() is helpful to identify the required spelling(s).

Here’s an example of how to get the population for the canton of Aargau:

library(propop) 
ag_population <- get_population(
  number_fso = "px-x-0102010000_101",
  year = 2023, 
  spatial_units = "- Aargau"
)

Get the parameters for a sample canton (mind using the same spelling as in the FSO tables; see comment above):

ag_parameters <- get_parameters(
  year = 2024,
  year_last = 2026,
  spatial_units = "Aargau"
)

The projection can be run as follows:

# select reference scenario
ag_parameters_ref <- ag_parameters |>
  dplyr::filter(scen == "reference")

propop(
  parameters = ag_parameters_ref,
  year = 2024,
  year_last = 2026,
  age_groups = 101,
  fert_first = 16,
  fert_last = 50,
  share_born_female = 100 / 205,
  population = ag_population,
  subregional = FALSE,
  binational = TRUE
)

Note of caution: As long as the FSO’s API interface and the underlying data structure remain stable, the functions will work. However, changes in the API are likely to break the functions.