Get Started • WorkInData

Cluster setup

The microdata are stored on Yale’s Bouchet HPC cluster.

Assuming you have a Yale identity, see the YCRC guide to get started on the cluster.

The package requires Arrow, which is available on the cluster in modules (e.g., Arrow/17).

Using the data

Start by installing the package if needed:

# install.packages("remotes")
remotes::install_github("dissc-yale/WorkInData")

Then load the library and connect to the data:

library(WorkInData)

data <- wid_open("/nfs/roberts/pi/dissc/work-in-data/gender_growth_gap")

Here, data is a regular Arrow Dataset, so you can use dplyr to interact with it, but the wid_subset function can be used to filter more efficiently.

For instance, we might look at working status among those with a graduate degree in 2018 by age and sex:

data_small <- data |>
  wid_subset(year == 2018, education == "Graduate") |>
  dplyr::select(age, sex, work) |>
  dplyr::collect()

data_small
#> # A tibble: 215,948 × 3
#>      age sex   work 
#>    <int> <lgl> <lgl>
#>  1    35 TRUE  TRUE 
#>  2    26 FALSE TRUE 
#>  3    36 TRUE  TRUE 
#>  4    65 TRUE  TRUE 
#>  5    70 FALSE TRUE 
#>  6    56 TRUE  FALSE
#>  7    41 FALSE TRUE 
#>  8    41 TRUE  TRUE 
#>  9    64 FALSE TRUE 
#> 10    24 TRUE  TRUE 
#> # ℹ 215,938 more rows

with(data_small, table(
  c("Yonger", "Older")[(age > mean(age, na.rm = TRUE)) + 1L],
  c("Male", "Female")[sex + 1L],
  work
))
#> , , work = FALSE
#> 
#>         
#>          Female  Male
#>   Older   18410 19166
#>   Yonger  17037  7536
#> 
#> , , work = TRUE
#> 
#>         
#>          Female  Male
#>   Older   30566 36366
#>   Yonger  45560 40699