Skip to contents

rmoriedata

Disclosure control for aggregate releases, with open datasets to use it on.

Install

Install the current release from r-universe (prebuilt binaries for macOS and Windows), which can be newer than CRAN. rmoriebricklayer, its companion, is named too, so an older copy already installed is replaced: rmoriedata needs rmoriebricklayer 1.1.1 or later.

With pak (progress bars, parallel downloads, compiler output hidden unless a build fails, and the named packages always upgraded to the current release). Keep the repos arguments as written: under Rscript there is no mirror chooser, so a bare install.packages() stops with “trying to use CRAN without setting a mirror”, and CRAN can carry older versions than r-universe.

if (!requireNamespace("pak", quietly = TRUE)) {
  install.packages("pak", repos = "https://cloud.r-project.org")
}
pak::repo_add(rootcoder007 = "https://rootcoder007.r-universe.dev")
pak::pkg_install(c("rmoriebricklayer", "rmoriedata"))

From a terminal (single quotes outside, double quotes inside, so the shell passes the R code through untouched):

Rscript -e 'if (!requireNamespace("pak", quietly = TRUE)) install.packages("pak", repos = "https://cloud.r-project.org"); pak::repo_add(rootcoder007 = "https://rootcoder007.r-universe.dev"); pak::pkg_install(c("rmoriebricklayer", "rmoriedata"))'

Without pak, plain install.packages() does the same with R’s own output:

Rscript -e 'install.packages(c("rmoriebricklayer", "rmoriedata"), repos = c("https://rootcoder007.r-universe.dev", "https://cloud.r-project.org"))'

Development version from GitHub (a source build; needs a C/C++ toolchain):

Rscript -e 'if (!requireNamespace("pak", quietly = TRUE)) install.packages("pak", repos = "https://cloud.r-project.org"); pak::repo_add(rootcoder007 = "https://rootcoder007.r-universe.dev"); pak::pkg_install("rootcoder007/rmoriedata")'
# without pak
Rscript -e 'repos <- c("https://rootcoder007.r-universe.dev", "https://cloud.r-project.org"); install.packages("remotes", repos = repos); remotes::install_github("rootcoder007/rmoriedata", repos = repos, upgrade = "always")'

On macOS, CRAN’s R (from https://CRAN.R-project.org/bin/macosx/) installs r-universe’s prebuilt binaries in seconds. Homebrew’s R cannot use them, so there every package is compiled from source.

Quick start 

library(rmoriedata)

# Release a private count, then check a table for re-identification risk.
dplapcnt(truecnt = 42, epsilon = 1.0)
df <- data.frame(age = c(25, 25, 25, 40), sex = c("F", "F", "F", "M"))
kanonvrf(df, c("age", "sex"), k = 2)$summary
cellsupp(matrix(c(120, 3, 47, 88), 2), threshold = 5)$protected

# Browse and load a bundled table by slug.
cat <- mdcat()
head(cat[cat$kind == "table", c("slug", "n_rows", "n_cols")])
iucr <- mdload("chicago_iucr_codes")

See vignette("rmoriedata") for the full tour.

Functions

Chicago sample data — R and Python

Bundled samples (complaint_sample, arrest_sample) from the City of Chicago open data; the full datasets are fetched on demand and cached.

library(rmoriedata)
data(complaint_sample)                       # bundled sample, offline
crimes <- chiload("complaints", full = TRUE, limit = 200)  # the full data, first 200 rows
# omit `limit` for the whole multi-million-row dataset (large: minutes and gigabytes)
pq <- chiload("arrests", as = "parquet_path")  # Parquet for Python
import pandas as pd
df = pd.read_parquet(pq)                      # path printed by chiload

# Or read the sample's .rda from the package source (data/complaint_sample.rda
# in the source tarball or the GitHub repository; an installed package keeps it
# in a lazy-load database instead) — use the date_iso column, since POSIXct
# does not roundtrip cleanly through pyreadr:
import pyreadr
complaints = pyreadr.read_r("data/complaint_sample.rda")["complaint_sample"]

Every bundled table also ships as Parquet: mdpath("<slug>") gives the verified path, pd.read_parquet() reads it.

Recommended Python bridge: the Parquet path (as = "parquet_path") — typed, columnar, read natively by pandas/polars/duckdb with correct timestamps and no R runtime. pyreadr on the source .rda works for quick read-only access but is slower and mangles POSIXct (use date_iso).

Curated tables at data.rmorie.com

The tables the MORIE project materialises from BigQuery public datasets, Statistics Canada and the survey, court and justice data its ledger uses are served from the edge with no sign-in: hostcat() lists them and hostload("db/table") opens one, checked against the SHA-256 its manifest lists; a large table comes down in parts that resume. Each table keeps its source’s licence; the terms are at www.rmorie.com/data-license/. ICPSR studies are not served there (ICPSR’s terms do not allow redistribution): hostcat() leaves them out and hostload() refuses them before any request. The endpoint comes from a signed services document the packages verify before use, so it can change without a release.

cat_ <- hostcat()            # key, name, rows, source dataset
df <- hostload("fec_cm_2020/fec_cm_2020")

Asking a language model

mdask() puts a question about the datasets to a language model you bring: a provider’s key (OpenAI, Anthropic, Google Gemini, OpenRouter, Groq, Mistral and more, each from its usual variable), an Ollama server, or an OpenAI-compatible server of your own. The model gets the catalogue and every function of rmoriedata and rmoriebricklayer, with tools to read their help.

Sys.setenv(ANTHROPIC_API_KEY = "...")          # or mdllmconf(anthropic.key = "...")
mdask("which tables cover police stops?")
mdask("which tables cover police stops?", route = "ollama", model = "qwen3:8b")

Citation

If you use rmoriedata in your research, please cite the software:

Ruhela, V. S. (2026). rmoriedata: Disclosure Control for Aggregate Releases, with Open Datasets. R package version 0.4.3. doi:10.32614/CRAN.package.rmoriedata. https://rootcoder007.r-universe.dev/rmoriedata

BibTeX (or run citation("rmoriedata") after installation for the entry stamped with the exact installed version, sourced from inst/CITATION):

@Manual{ruhela_rmoriedata_2026,
  title  = {rmoriedata: Disclosure Control for Aggregate Releases, with Open Datasets},
  author = {Ruhela, Vansh Singh},
  year   = {2026},
  doi    = {10.32614/CRAN.package.rmoriedata},
  url    = {https://rootcoder007.r-universe.dev/rmoriedata}
}

See CITATION.cff for the machine-readable metadata GitHub’s “Cite this repository” button uses.

License

AGPL-3.0-or-later for the code. The bundled data are snapshots of public data, each under the licence or terms of use of the body that published it; the publisher and source of every table are named in its help page (?complaint_sample, ?arrest_sample, ?siuload, ?cihitbls, ?`rmoriedata-victoria`), in the catalogue (mdcat(), from inst/extdata/_catalog.csv), and in the data-attribution section of inst/NOTICE.