Skip to content

fix: 🐛 use duckplyr to read Parquet as DuckDB #348

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Aastedet wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: 🐛 use duckplyr to read Parquet as DuckDB #348

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion DESCRIPTION
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@ BugReports: https://github.com/dp-next/fastreg/issues
Depends:
R (>= 4.1.0)
Imports:
duckplyr (>= 1.2.1),
arrow,
checkmate,
cli,
Expand All @@ -46,7 +47,8 @@ Suggests:
testthat (>= 3.0.0),
tidyselect,
withr,
tarchetypes
tarchetypes,
duckdb (>= 1.5.4)
VignetteBuilder:
quarto
Encoding: UTF-8
Expand Down
21 changes: 13 additions & 8 deletions R/read.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -57,13 +57,15 @@ read_parquet_dataset <- function(path) {
checkmate::assert_directory_exists(path)
assert_directory_not_empty(path)
path |>
arrow::open_dataset(
unify_schemas = TRUE,
# Explicitly set type of partition to int32 to handle when year is
# missing.
partitioning = arrow::hive_partition(year = arrow::int32())
duckplyr::read_parquet_duckdb(
prudence = "thrifty",
options = list(
union_by_name = TRUE,
hive_partitioning = TRUE
)
) |>
arrow::to_duckdb()
duckplyr::as_duckdb_tibble() |>
duckplyr::as_tbl()
Comment on lines +67 to +68

@lwjohnst86 lwjohnst86 Jul 17, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why are both these needed? And doesn't that negate the thrifty?

@Aastedet Aastedet Jul 18, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I’ll have to look into that. It mirrors the existing osdc testing pipeline when we set up the simulated data, so I assumed it was good. I didn’t dive deep into the technical details, this was just what “worked”, so I ran with it.

The prudence choice is also quite important - I figured thrifty is going to strike the best balance between memory safety and practical usefulness for quick exploratory summary statistics etc, but it’s definitely up for discussion.

}

#' @describeIn read_parquet Reads a single Parquet file.
Expand All @@ -73,8 +75,11 @@ read_parquet_file <- function(path) {
checkmate::assert_file_exists(path)
assert_is_parquet(path)
path |>
arrow::read_parquet() |>
arrow::to_duckdb()
duckplyr::read_parquet_duckdb(
prudence = "thrifty"
) |>
duckplyr::as_duckdb_tibble() |>
duckplyr::as_tbl()
}

assert_is_parquet <- function(path) {
Expand Down
6 changes: 4 additions & 2 deletions README.qmd
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,9 +10,11 @@ gh:
<!-- badges: start -->

[![CRAN status](https://www.r-pkg.org/badges/version/fastreg)](https://CRAN.R-project.org/package=fastreg)
[![GitHub Release](https://img.shields.io/github/v/release/{{< meta gh.org >}}/{{< meta gh.repo >}}.svg)](https://github.com/{{< meta gh.org >}}/{{< meta gh.repo >}}/releases/latest)
[![GitHub
Release](https://img.shields.io/github/v/release/{{< meta gh.org >}}/{{< meta gh.repo >}}.svg)](https://github.com/{{< meta gh.org >}}/{{< meta gh.repo >}}/releases/latest)
[![Build](https://github.com/{{< meta gh.org >}}/{{< meta gh.repo >}}/actions/workflows/build.yml/badge.svg)](https://github.com/{{< meta gh.org >}}/{{< meta gh.repo >}}/actions/workflows/build.yml)
[![pre-commit.ci status](https://results.pre-commit.ci/badge/github/{{< meta gh.org >}}/{{< meta gh.repo >}}/main.svg)](https://results.pre-commit.ci/latest/github/{{< meta gh.org >}}/{{< meta gh.repo >}}/main)
[![pre-commit.ci
status](https://results.pre-commit.ci/badge/github/{{< meta gh.org >}}/{{< meta gh.repo >}}/main.svg)](https://results.pre-commit.ci/latest/github/{{< meta gh.org >}}/{{< meta gh.repo >}}/main)
[![lifecycle](https://lifecycle.r-lib.org/articles/figures/lifecycle-experimental.svg)](https://lifecycle.r-lib.org/articles/stages.html#experimental)
[![Project Status: Active – The project has reached a stable, usable state and is being actively developed.](https://www.repostatus.org/badges/latest/active.svg)](https://www.repostatus.org/#active)

Expand Down
37 changes: 10 additions & 27 deletions tests/testthat/test-read.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,8 @@ purrr::walk(sas_paths$output_path, \(path) {
# Test read_parquet_*() ---------------------------------------------------------

test_that("read_parquet_file() reads a single Parquet file", {
gc() # Force any previous connection to be dropped to avoid "View with name X already exists"-errors during tests

# Read single Parquet file (2020 file).
# Because UUID is used in the convert function, we can't know the name of the
# file.
Expand All @@ -27,8 +29,8 @@ test_that("read_parquet_file() reads a single Parquet file", {
expected_data <- haven::read_sas(expected_source_file)

expect_equal(
# year col doesn't exist when only one file is read.
actual_data |> dplyr::select(-"source_file"),
# source file and year columns don't exist in the basic haven conversion
actual_data |> dplyr::select(-c("source_file", "year")),
expected_data,
ignore_attr = TRUE
)
Expand All @@ -39,6 +41,7 @@ test_that("read_parquet_file() reads a single Parquet file", {
})

test_that("read_parquet_dataset() reads a partitioned Parquet register", {
gc() # Force any previous connection to be dropped to avoid "View with name X already exists"-errors during tests
actual <- read_parquet_dataset(output_dir) |> dplyr::collect()

expected <- purrr::map(sas_paths$output_path, \(path) {
Expand Down Expand Up @@ -106,6 +109,7 @@ test_that("read_parquet_file() errors when file is not Parquet", {
})

test_that("files with extension .parq can also be read", {
gc() # Force any previous connection to be dropped to avoid "View with name X already exists"-errors during tests
path <- fs::path_temp("file.parq")
arrow::write_parquet(
simulate_registers_with_paths("bef")$data[[1]],
Expand All @@ -114,8 +118,9 @@ test_that("files with extension .parq can also be read", {
expect_no_error(read_parquet_file(path))
})


test_that("read_parquet_dataset() reads files with different columns", {
gc() # Force any previous connection to be dropped to avoid "View with name X already exists"-errors during tests

# Faux bef with lmdb structure, saved separately and combined with sas paths
sas_dir <- fs::path_temp("different_columns/sas")
parquet_dir <- fs::path_temp("different_columns/parquet")
Expand Down Expand Up @@ -147,33 +152,11 @@ test_that("read_parquet_dataset() reads files with different columns", {
)
})

test_that("read_parquet_dataset() errors with incompatible schemas", {
# Create a bef file where numeric columns are changed to character, so
# the schema is incompatible with the other bef files.
incompatible_data <- bef_list$data[[1]] |>
dplyr::mutate(dplyr::across(where(is.numeric), as.character))

incompatible_sas_path <- fs::path_temp(
"sas_schema_incompatible/bef2099.sas7bdat"
)
write_to_sas(incompatible_data, incompatible_sas_path)
sas_incompatible <- c(
sas_paths$output_path,
incompatible_sas_path
)

incompatible_output <- fs::path_temp("incompatible")
# Convert files.
purrr::walk(sas_incompatible, \(path) {
convert(path, incompatible_output)
})

expect_error(read_parquet_dataset(incompatible_output), "incompatible")
})

Comment on lines -150 to -173

@lwjohnst86 lwjohnst86 Jul 17, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why remove this? The point was to ensure that it correctly fails if there is an incompatible schema.

@Aastedet Aastedet Jul 18, 2026

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The duckdb reader is inherently much more robust to type mismatches than Arrow’s. There doesn’t appear to be a way to make duckplyr::read_parquet_duckdb() fail that test, so I think it has to be removed.

# Test read_register() ---------------------------------------------------------

test_that("read in a register", {
gc() # Force any previous connection to be dropped to avoid "View with name X already exists"-errors during tests

withr::with_options(
list(
fastreg.project_rawdata_dir = fs::path_temp("E/rawdata/202020/"),
Expand Down
26 changes: 22 additions & 4 deletions tests/testthat/test-use.R
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -91,17 +91,35 @@ test_that("targets pipeline template converts SAS files to Parquet", {
purrr::list_c() |>
sum()

n_actual_bef <- arrow::open_dataset(fs::path(
n_actual_bef <- fs::path(
test_output_dir,
"bef"
)) |>
) |>
duckplyr::read_parquet_duckdb(
prudence = "thrifty",
options = list(
union_by_name = TRUE,
hive_partitioning = TRUE
)
) |>
duckplyr::as_duckdb_tibble() |>
duckplyr::as_tbl() |>
dplyr::collect() |>
nrow()

n_actual_lmdb <- arrow::open_dataset(fs::path(
n_actual_lmdb <- fs::path(
test_output_dir,
"lmdb"
)) |>
) |>
duckplyr::read_parquet_duckdb(
prudence = "thrifty",
options = list(
union_by_name = TRUE,
hive_partitioning = TRUE
)
) |>
duckplyr::as_duckdb_tibble() |>
duckplyr::as_tbl() |>
dplyr::collect() |>
nrow()

Expand Down
Loading