CleanLASA

CleanLASA helps you work with LASA (Longitudinal Aging Study Amsterdam) data in R. LASA data comes as SPSS (.sav) files, and each file uses its own short-hand variable names and numeric codes, which change slightly from one measurement wave to the next. Reading such a file with a generic tool like haven::read_sav() gives you a data frame of bare numbers with no explanation of what they mean.

CleanLASA fixes that. Point it at a LASA .sav file and it:

• figures out which LASA file and measurement wave you’re looking at, from the file name;
• attaches the correct variable and value labels for that specific file/wave, so 4 reads as "married" instead of just 4;
• lines up the same variable across waves, even when its name or its numeric codes changed between waves;
• tells you plainly which variables it could not figure out, so you can check them by hand instead of silently getting it wrong; and
• helps you find the right LASA topic, file code, and official documentation in the first place.

All of this metadata lives in one bundled database (lasa_label_db()), hand-transcribed from LASA’s own documentation and shipped with the package. It gets more complete and more accurate with every package update, and you can also patch your own local copy with manual_update_lasa_labels() if something needs a quick, local fix.

Installation

CleanLASA needs R 4.0 or later. Install the latest release (v1.0.0) from GitHub with:

# install.packages("remotes")
remotes::install_github("Highmeadows/CleanLASA@v1.0.0")

or the development version, with changes that haven’t been released yet, with:

remotes::install_github("Highmeadows/CleanLASA")

See NEWS.md for what changed in each release.

Quick start

The main function is read_lasa_sav(). Give it the path to a LASA .sav file and it works out the rest from the file name – reading the file, matching its columns to LASA’s documentation, attaching labels, turning categorical variables into R factors, and renaming columns to a name that’s the same across every wave:

library(CleanLASA)

data <- read_lasa_sav("LASAB046.sav") # LASA wave B, file code 046

# Check whether anything didn't match:
lasa_label_report(data, problems_only = TRUE)

Z files such as LASAZOA1.sav store several waves side by side (boak, coak, doak, …). With format_as_long = TRUE you get them in long format instead: one row per respondent per wave, with a Wave column, a Time column counting the waves since the respondent’s baseline, and one oak column. transform_lasa_data() converts any LASA data – including merged files – between long and wide format:

oa <- read_lasa_sav("LASAZOA1.sav", format_as_long = TRUE)
oa_wide <- transform_lasa_data(oa, format = "wide") # boak, coak, ...

Not sure which file code you need? lasa_topics() searches LASA’s topics and themes for you, and lasa_var_info() opens LASA’s own documentation for a topic or file code:

lasa_topics(topic = "physical activity")
lasa_var_info("046")

Learn more

This README is just a summary. The package vignette is a full, step-by-step tutorial that walks through every function with worked examples – how the file-name-based matching works, every option read_lasa_sav() offers (fuzzy matching, name corrections, converting to factors/numeric, standardizing names and labels across waves), what’s inside the label database, how to record your own corrections, how to reshape data between long and wide format, and how to search LASA’s topic index and documentation. Start there if you’re new to the package.

Functions

Function	Description
read_lasa_sav()	Reads a LASA .sav file and labels it using lasa_label_db(), based on the wave/file code parsed from the file name
apply_lasa_labels()	Applies lasa_label_db() labels to any data frame, resolving file code/wave explicitly or from stored provenance
transform_lasa_data()	Transforms LASA data between wide format (one row per respondent, as in Z files) and long format (one row per respondent per wave, with Wave and Time columns)
manual_update_lasa_labels()	Hand-corrects or adds a variable/value label in your local copy of the database
restore_lasa_labels()	Undoes manual corrections (entirely, or scoped) and/or rebuilds the local database cache from the bundled one
lasa_label_db()	Returns the label database currently in effect (bundled snapshot + any local updates)
lasa_label_report()	Returns variables that could not be matched to the corresponding LASA documentation
lasa_topics()	Searches the bundled LASA topic database for topics, themes, and file codes
lasa_var_info()	Finds and opens a topic’s or file code’s variable-information PDF or LASA webpage, resolved live from the LASA website

Documentation

Full function reference and articles are available at: https://highmeadows.github.io/CleanLASA/