Advanced Querying"
In nomisdata: Access 'Nomis' UK Labour Market Data and Statistics

knitr::opts_chunk$set(collapse = TRUE, comment = "#>", eval = FALSE)
# Use temp directory during vignette build to avoid CRAN NOTE
Sys.setenv(NOMISDATA_CACHE_DIR = file.path(tempdir(), "nomisdata"))

Large Data Downloads

When downloading large datasets, nomisdata automatically handles pagination:

library(nomisdata)

# This will automatically chunk if >25,000 rows (guest) or >100,000 rows (API key)
large_data <- fetch_nomis(
  "NM_1_1",
  time = c("first", "latest"),  # All time periods
  geography = "TYPE464",         # All local authorities
  measures = 20100
)

Caching for Reproducibility

Enable caching to avoid re-downloading data:

# Enable caching
enable_cache(file.path(tempdir(), "nomis_cache"))

# First call downloads from API
data1 <- fetch_nomis("NM_1_1", time = "latest", geography = "TYPE499")

# Second call uses cache
data2 <- fetch_nomis("NM_1_1", time = "latest", geography = "TYPE499")

# Clear when needed
clear_cache()

Parallel Downloads

For very large queries across multiple parameters:

library(future)
plan(multisession, workers = 4)

# Define queries
queries <- list(
  list(geography = "2092957697", time = "2020"),
  list(geography = "2092957698", time = "2020"),
  list(geography = "2092957699", time = "2020")
)

# Fetch in parallel (implementation would use future_map)

Working with Multiple Datasets

# Employment data
employment <- fetch_nomis(
  "NM_168_1",
  time = "latest",
  geography = "TYPE499"
)

# Benefits data
benefits <- fetch_nomis(
  "NM_1_1",
  time = "latest",
  geography = "TYPE499"
)

# Join by geography
library(dplyr)
combined <- employment |>
  inner_join(benefits, by = "GEOGRAPHY_CODE", suffix = c("_emp", "_ben"))