project.pasfr: Projections of Percent Age-Specific Fertily Rate

In bayesPop: Probabilistic Population Projection

Projections of Percent Age-Specific Fertily Rate

Description

The projections of percent age-specific fertility rate (PASFR) is normally computed within the pop.predict function for each trajectory. This function allows to project PASFR outside of population projections for the median total fertility rate (TFR) or user-provided TFR, and export it.

Usage

project.pasfr(inputs = NULL, present.year = 2020, end.year = 2100, 
    wpp.year = 2019, annual = FALSE, nr.est.points = if(annual) 15 else 3,
    digits = 2, out.file.name = "percentASFR.txt", verbose = FALSE)
    
project.pasfr.traj(inputs = NULL, countries = NULL, nr.traj = NULL, 
    present.year = 2020, end.year = 2100, wpp.year = 2019, 
    annual = FALSE, nr.est.points = if(annual) 15 else 3,
    digits = 2, out.file.name = "percentASFRtraj.txt", verbose = FALSE)

Arguments

inputs	List of input data (file names) with the same meaning as in pop.predict. The relevant items here are: either tfr.file or tfr.sim.dir (TFR estimates and projections), pasfr (PASFR for observed time periods), and patterns (PASFR patterns). All entries are optional. By default the data is taken from the corresponding wpp package. See Details below.
present.year	Year of the last observed data point.
end.year	End year of the projection.
wpp.year	Year for which WPP data is used if one of the inputs components is left out.
annual	Logical that should be TRUE if the provided data on TFR and PASFR are annual-based data.
nr.est.points	Number of time points to be used for estimating the continuation of the observed PASFR trend. By default it is 15 years, corresponding to three time points for 5-year data.
digits	Number of decimal places in the results.
out.file.name	Name of the resulting file. If NULL nothing is written.
verbose	Logical switching verbose messages on and off.
countries	Vector of numerical country codes. By default the function is applied to all countries.
nr.traj	Number of trajectories on which the function should be applied. By default all trajectories are taken. Otherwise they are thinned appropriately.

Details

If the input TFR is given as an ASCII file (in inputs$tfr.file), it can be either a csv (comma-separated) file in long format, with columns “LocID”, “Year”, “Trajectory” and “TF”. Or, it can be a tab-separated (wide format) file with column “country_code” and each year or time period as a separate column (see tfr). In the latter case, an additional inputs entry tfr.file.type = "w" must be provided to specify the file is in the wide format, which is a case whe there is only one trajectory. Note that the TFR input should cover all projection time period as well as observed TFR as the function assesses the start of Phase III, which could be in the past.

If observed PASFR is given (in inputs$pasfr), it is a tab-separated file in wide format as in percentASFR. Fertility age patterns can be controlled by country via the inputs$patterns entry, which is a dataset in the same format and meaning as vwBaseYear.

In addition, if the present year differs by country, the inputs list accepts the entry last.observed, which is a tab-separated file with columns “country_code” and “last.observed”. It can contain the year of the last observed time period for each country.

In the project.pasfr function, if the TFR input (given either as a long file or as a simulation directory), contains more than one trajectory, the median is derived over the trajectories for each time period. Then, PASFR corresponding to this median is projected using the method from Sevcikova et al (2015).

For project.pasfr.traj, the PASFR is projected for single trajectories of TFR.

Value

Returns invisible data frame with the projected PASFR.

Author(s)

Hana Sevcikova, Igor Ribeiro

References

H. Sevcikova, N. Li, V. Kantorova, P. Gerland and A. E. Raftery (2016). Age-Specific Mortality and Fertility Rates for Probabilistic Population Projections. In: Dynamic Demographic Analysis, ed. Schoen R. (Springer), pp. 285-310. Earlier version: https://arxiv.org/abs/1503.05215

See Also

pop.predict

Examples

# using TFR in simulation directory
inputs <- list(tfr.sim.dir=file.path(find.package("bayesTFR"), "ex-data", "bayesTFR.output"))
pasfr <- project.pasfr(inputs, out.file.name = NULL)
head(pasfr)

## Not run: 
pasfr.traj <- project.pasfr.traj(inputs, out.file.name = NULL)
head(pasfr.traj)
## End(Not run)

# using TFR in wide-format file
inputs2 <- list(tfr.file = file.path(find.package("wpp2019"), "data", "tfrprojMed.txt"),
    tfr.file.type = "w")
pasfr2 <- project.pasfr(inputs2, out.file.name = NULL)
head(pasfr2)

bayesPop documentation built on Aug. 10, 2023, 1:10 a.m.