VirtualPop

$VirtualPop$ generates a virtual population from demographic data. The demographic data are death rates (mortality rates) by age and sex, and birth rates (fertility rates) by age and birth order (parity). The current version of VirtualPop uses rates downloaded from the Human Mortality Database (HMD) and the Human Fertility Database (HFD).

$VirtualPop$ simulates, for each member of a virtual population, the lifespan and the fertility history. Simulation is in continuous time. The simulation model is an individual-level state transition (multi-state) model. Simulated life histories are stored in a data structure commonly used for sample surveys. The data structure is a data frame (flat file) with one record per individual. Personal attributes are represented as column variables. If requested, $VirtualPop$ generates a multi-generation virtual population by simulating life (fertility) histories of children, grandchildren, and great-grandchildren. The genealogies facilitate the study of family ties and kinship networks implicit in a set of demographic rates.

The package comes with four vignettes.

• VirtualPop: Simulation of individual fertility careers from age-specific fertility and mortality rates. A tutorial. A good place to start.
• Sampling piecewise-exponential waiting time distributions. The theory underlying the sampling of ages at death and ages at childbirth.
• Simulation of life histories. The theory underlying the generation of life histories using multistate models.
• Validation of the simulation. Results of the simulation are compared with demographic (kinship) indicators computed from census and survey data. The virtual population generated from demographic rates of the United States in 2019 is compared with observations in the Current Population Survey 2018.

In the $doc$ folder, you find

• pdf versions of the vignettes
• package manual
• R code used in vignettes

The companion package $Families$ extract family relationships from the multi-generation virtual population. These relationships are the basis for the computation of kinship indicators. Areas of application include kinship networks, the demography of grandparenthood, the demography of sandwich generations (double burden of child care and parental care), and perspectives of children on population.

To install the package from CRAN, type in the R window:

    install.packages ("VirtualPop")

To install the package from GitHub, use:

    library(devtools)
    devtools::install_github("willekens/VirtualPop")