Installation

You can install the development version of tidylfs from GitHub with:

# install.packages("devtools")
devtools::install_github("wklimowicz/tidylfs")

Quickstart

Copy the LFS .sav files into a directory (in my example lfs_data_folder). They should look like this:

.
└── lfs_data_folder/
    ├── 1992 Q1.sav
    ├── 1992 Q2.sav
    ├── ...
    └── 2020 Q2.sav

You can then run:

library(tidylfs)

# Converts `.sav` files to `.Rds` files, to save
# space and for quicker loading
lfs_convert(lfs_directory = "lfs_data_folder/",
            output_directory = "lfs_rds_folder/")

# Compiles into one file.
lfs <- lfs_compile(lfs_directory = "lfs_rds_folder/")

To reproduce official ONS publications, such as

Unemployent UNEM01 NSA

lfs %>%
    lfs_summarise_unemployment(QUARTER)

Salary by occupation EARN06

lfs %>%
    dplyr::filter(
      !is.na(OCCUPATION_MAJOR), # Filter out NA's
      FTPTWK == "Full-time" # Take only full-time employees
    ) %>%
    lfs_summarise_salary(QUARTER, OCCUPATION_MAJOR)

Extending them is easy:

Unemployment by quarter, sex, and age category

lfs %>%
  lfs_summarise_unemployment(QUARTER, SEX, AGES)

For more information on the ONS variables, see the LFS Guidance.

Some of the variables included by default are:

Variable Name	Definition
YEAR	Year
QUARTER	Quarter
SEX	Sex
GOVTOR	Government Office Region
AGE	Age
FTPTWK	Part-Time/Full-time Status
EDAGE	Age when completed full time education
TTACHR	Actual hours worked
UNION	In union?
HIQUALD	Highest Qualification
DEGREE_SUBJECT	Degree Subject
OCCUPATION	Occupation in main job
PARENTAL_OCCUPATION	Parental Occupation at 14
ETHNICITY	Ethnicity

Using the data across multiple projects

To avoid storing multiple copies of the compiled dataset, you can set the DATA_DIRECTORY environment variable before compiling. This can be done on the system or in .Rprofile:

Sys.setenv(DATA_DIRECTORY = "path/to/folder")

When compiling the dataset include the save_to_folder option as TRUE:

lfs_compile(lfs_directory = "lfs_rds_folder/", save_to_folder = TRUE)

The compiled dataset will be saved to the directory as an fst file, and you can load it from anywhere using this command:

lfs <- lfs_load(data.table = FALSE)