Install RegiStream

Stata

Requires Stata 16.0 or later.

Install everything (recommended)

The registream meta-package installs the shared core + every module in one command:

net install registream, from("https://registream.org/install/stata/latest") replace

After this you have registream (core: configuration, first-run wizard, update management, shared utilities), autolabel, and datamirror — no further installs needed.

Install a single module (alternative)

If you only want one module, install just that one. Each module pulls in registream core automatically as a dependency, so you don't need to install core separately either way.

* Just autolabel + core (no datamirror)
net install autolabel,  from("https://registream.org/install/stata/latest") replace

* Just datamirror + core (no autolabel)
net install datamirror, from("https://registream.org/install/stata/latest") replace

Python

Requires Python 3.11 or later.

From PyPI

pip install registream
pip install registream-autolabel

The registream meta-package installs the shared core. registream-autolabel adds the autolabel module on top.

API surface: df.rs.autolabel(), scope(), suggest(), df.rs.lookup(), info(), cite(), with pandas accessor + shortcut monkey-patches and matplotlib/seaborn label integration. See the Python reference for full signatures.

R

Requires R 4.1 or later.

Direct install (current)

install.packages(
  c("registream", "autolabel"),
  repos = c("https://registream.org/r/",
            "https://cloud.r-project.org/"),
  type = "source"
)

The second repos entry is the CRAN fallback for transitive deps (curl, digest, jsonlite, haven). Source install only, no compiled code. CRAN submission is on the way; once accepted, install.packages(c("registream", "autolabel")) from the default CRAN mirror will work too.

API surface: autolabel(), scope(), suggest(), rs_lookup(), rs_lab(), info(), cite(). See the R reference for full signatures.

First run

The first time you call an autolabel command, a short setup wizard picks a mode:

• Offline — no network, no telemetry, no auto-update checks. For air-gapped environments (MONA, Forskermaskinen, Dapla).
• Standard — network allowed (update checks), no telemetry. Recommended for most users.
• Full — network allowed, batched usage telemetry on to help improve RegiStream.

On R only, a second prompt asks where to write config + cache — ~/.registream/ (shared with Stata and Python) or R's isolated tools::R_user_dir("registream","cache"). The extra prompt is a CRAN-compliance requirement; Stata and Python always use ~/.registream/ (or $REGISTREAM_DIR if set). You only see the wizard once.

use my_register_data.dta, clear
autolabel variables, domain(scb) lang(eng)
* → setup wizard runs once, then labels are applied

Secure environments

Register-data research often happens in secure environments (MONA for SCB, Forskermaskinen for DST, Dapla for SSB, and others). These have no internet access, so autolabel update datasets can't reach registream.org. The workflow is: download on a dev machine → transfer in via the agency's file-in channel → install offline.

Full walkthrough (with per-environment notes for MONA, Forskermaskinen, and Dapla): Secure-environment install →

Institutional setup

Deploying RegiStream to a research group, a department, or a secure-env project tenant? Shared configuration + private domains let you run autolabel variables, domain(hospital) on in-house register data with no internet and no coordination with registream.org.

Full walkthrough (shared config, private-domain construction, hybrid deployments): Institutional setup →

Verify install

* Stata
registream version
registream info

# Python
python -m registream version
python -m registream.autolabel version

# R
library(autolabel); info()

Each should print the installed core + module versions, the cache location, and the currently configured domains.