Skip to contents

Prepare Data from Human Internal Migration Database

data_ssvd_himd.Rd

Process data on age-specific migration probabilities from the Human Internal Migration Database.

Usage

data_ssvd_himd(
  zipfile,
  time_interval = 1,
  measure_type = c("rate", "prob"),
  n_comp = 5,
  eps = 1e-05
)

Arguments

zipfile

The name of a zipped file downloaded from the Human Internal Migration Database. A path name that is handled by utils::unzip().

time_interval

Length of interval over which values are calculated. Choices are 1 or 5. Default is 1.

measure_type

"prob" or "rate". Whether values are to be treated as probabilities or as rates. See data_ssvd_himd() for details. Default is "rate".

n_comp

Number of SVD components to include in result. Default is 5.

eps

Parameter for truncating probabililities or rates. Default is 0.00001.

Value

A tibble

Usage

Step 1: Download data

Download the data from Dyrting, S. (2024, October 23). Data from: Estimating Complete Migration Probabilities from Grouped Data at https://osf.io/vmrfk/. The data is in the "Migration" folder under the "Files" tab. The facility for downloading all the files at once does not appear to be working, but CSV files can be downloaded one by one. Zip the folder containing these CSV files.

**Step 2: Call function data_ssvd_himd()

Create three datasets:

himd_data_rate <- data_svd_himd(zipfile = "himd.zip",
                                time_interval = 1,
                                measure_type = "rate")
himd_data_prob1 <- data_svd_himd(zipfile = "himd.zip",
                                 time_interval = 1,
                                 measure_type = "prob")
himd_data_prob5 <- data_svd_himd(zipfile = "himd.zip",
                                 time_interval = 5,
                                 measure_type = "prob")

Rates or probabilities

The original data are probabilities. However, data_ssvd_himd() (and coef_himd()) allow values where time_interval is 1 to be to treated as rates. If more than one change in residence in a year is unusual, then one-year probabilities and should in fact be very similar.

coef_himd Truncation and transformation

See also

  • coef_hfd() Obtain time series of SVD coefficients for Human Internal Migration Database data

  • tidy_hmd() Format data from the Human Internal Migration Database into a tidy data frame

Examples

zipfile <- system.file("extdata", "himd_20241023_subset.zip",
                       package = "bssvd")
data_ssvd_himd(zipfile)
#> Unzipping file...
#> Creating five-year age groups...
#> Assembling datasets for alternative open age groups...
#> Carrying out SVD...
#> # A tibble: 22 × 5
#>    type  labels_age  labels_sexgender matrix         offset     
#>    <chr> <list>      <list>           <list>         <list>     
#>  1 total <chr [61]>  <NULL>           <dgCMatrx[,5]> <dbl [61]> 
#>  2 total <chr [66]>  <NULL>           <dgCMatrx[,5]> <dbl [66]> 
#>  3 total <chr [71]>  <NULL>           <dgCMatrx[,5]> <dbl [71]> 
#>  4 total <chr [76]>  <NULL>           <dgCMatrx[,5]> <dbl [76]> 
#>  5 total <chr [81]>  <NULL>           <dgCMatrx[,5]> <dbl [81]> 
#>  6 total <chr [86]>  <NULL>           <dgCMatrx[,5]> <dbl [86]> 
#>  7 total <chr [91]>  <NULL>           <dgCMatrx[,5]> <dbl [91]> 
#>  8 total <chr [96]>  <NULL>           <dgCMatrx[,5]> <dbl [96]> 
#>  9 total <chr [101]> <NULL>           <dgCMatrx[,5]> <dbl [101]>
#> 10 total <chr [106]> <NULL>           <dgCMatrx[,5]> <dbl [106]>
#> # ℹ 12 more rows