Package 'datadict'

Title: Validate Data Against a 'data-dict.yaml' Data Dictionary
Description: A data dictionary describes a collection of related tables: their columns, types, constraints, relationships, and glossary. Validate a dataset against its dictionary and read the findings as an HTML report. Reference: <https://data-dict.tidyverse.org>.
Authors: Gábor Csárdi [aut, cre] (ORCID: <https://orcid.org/0000-0001-7098-9676>), Hadley Wickham [aut] (ORCID: <https://orcid.org/0000-0003-4757-117X>), Posit Software, PBC [cph, fnd] (ROR: <https://ror.org/03wc8by49>)
Maintainer: Gábor Csárdi <[email protected]>
License: MIT + file LICENSE
Version: 0.1.0
Built: 2026-09-18 11:37:46 UTC
Source: https://github.com/tidyverse/data-dict

Help Index

Download the data-dict binary

Description

Downloads the release archive for this platform, checks it against the published SHA-256, and unpacks the binary into tools::R_user_dir("datadict", "cache").

Usage

dd_install(version = "latest", force = FALSE, quiet = FALSE, dest = NULL)

Arguments

version

Release to install, e.g. "0.0.1" (a leading v is also accepted). Defaults to the latest release.
force

Whether to download again when a binary is already installed.
quiet

Whether to suppress the download progress bar and messages.
dest

The destination path for the downloaded binary. Defaults to machines canonical data directory, see tools::R_user_dir(). Warning: if you change the default then datadict will not be able to find it automatically, so we strongly suggest leaving it as the default.

Value

The path to the installed binary, invisibly.

Examples

dd_install()

Locate the data-dict binary

Description

Searched in order: the DATA_DICT environment variable, the copy installed by dd_install() in the package's cache directory, then the PATH.

Usage

dd_path(check = TRUE)

Arguments

check

Whether to throw an error when no binary is found. With FALSE, return "" instead.

Value

Path to the binary, or "" if it was not found and check is FALSE.

Examples

dd_path(check = FALSE)

Run the data-dict binary

Description

Both output streams are captured, interleaved as the binary wrote them. With echo = TRUE they are also printed as they arrive, so a long run shows its progress.

Usage

dd_run(args, echo = FALSE, ...)

Arguments

args

Character vector of command line arguments.
echo

Whether to print the binary's output while it runs, on top of capturing it.
...

Passed on to processx::run(), e.g. timeout or wd.

Details

A non-zero exit status is an error, carrying the captured output as the message.

Value

A list with the exit status and the captured output, a character vector of lines, invisibly.

Examples

dd_run("--version")

Validate a dataset's values against its data dictionary

Description

Runs ⁠data-dict render-report⁠, which validates the data and writes the findings as a self-contained HTML report, and opens it in a browser. The data files themselves come from each table's ⁠source: parquet:⁠ entry in the dictionary, resolved relative to the dictionary's directory.

Usage

dd_validate_data(
  dict = ".",
  table = NULL,
  html = tempfile(fileext = ".html"),
  browse = interactive()
)

Arguments

dict

A data-dict.yaml file, or a directory holding one.
table

Validate only this table, instead of every table in the dictionary.
html

Where to write the report.
browse

Whether to open the report with utils::browseURL().

Details

A dataset that fails validation is not an R error: the report is the point, so check status to see whether the run passed.

Value

A list with the html path and the CLI's exit status, where 0 means the dataset validated, invisibly.

Examples

# a directory holding a data-dict.yaml and the parquet files it points at
dd_validate_data("inst/data")