MORIE 森

Multi-domain Open Research and Inferential Estimation.

A multi-domain scientific computing toolkit (Python and R) for observational inference, with sociolegal, signal-processing, cryptographic, spatial-statistics, statistical-physics, and psychometrics modules. Hosts the MRM framework as a primary application for Canadian carceral, police, and oversight data analysis.

The R package ships native causal-inference engines — matching (nearest/Mahalanobis/exact/CEM/optimal/genetic/cardinality), double machine learning with clustered SEs, R-learner causal forests, T/S/X/DR meta-learners, design-based GLM, and a causal DAG toolkit (morie_dag*) — implemented in-package and cross-validated against MatchIt, DoubleML, grf, and dagitty rather than depending on them.

The morie command line checks PyPI once a day for a newer release (fail-silent, cached); import morie makes no network request. Set MORIE_NO_UPDATE_CHECK=1 to disable it.

Installation

Full step-by-step install guide with platform-specific notes (PEP 668 on Debian, python 3.13 segfault on Raspberry Pi OS, etc.) is at INSTALLATION.md.

morie is a Python (and R) package — once Python is present it is pip install morie. If you are starting with nothing installed, INSTALLATION.md opens with Step 1 — install the prerequisites: every tool you might need (Python, curl, bash/WSL, Git Bash, winget, Homebrew, Docker, R) with its official download. The short version:

Two steps, whatever the channel. Install morie, then run

morie interactive install

once per user. The published package leaves out the five modules behind morie repl, morie exec, morie agent and morie tui (they run code you or a model type); that command fetches them for your installed version from the release tag on GitHub and checks each against the manifest inside the package. The one-liner installer below runs it for you. A verb that needs the layer says so and, on a terminal, offers to run it. Details: the interactive layer.

• Windows — install Python from python.org (on the first screen tick Add python.exe to PATH), then pip install morie. Full walkthrough: Windows below. Windows has no curl/bash, so the one-liner does not apply there.
• macOS / Linux — the one-liner below sets up everything. It needs curl and bash, which macOS has built in and most Linux ships.
• Already have Python ≥3.10 — just pip install morie (on Debian, Ubuntu and Raspberry Pi OS use the venv three-liner below; their system pip refuses to install into /usr). The estimators take a pandas DataFrame, a CSV path, or a dict of columns; pandas is optional (morie ships its own frame core).

For terminal users — one-liner (Linux / macOS / WSL)

The simplest path if you have a terminal with curl and bash — both are built into macOS and preinstalled on most Linux (Windows has no bash, so use the installer above instead). It then bootstraps everything else for you: Python via uv, a managed venv, and the morie wheel. No pre-existing Python or pip needed.

curl -fsSL https://rootcoder007.github.io/morie/install.sh | bash

Or, with R alongside Python:

curl -fsSL https://rootcoder007.github.io/morie/install.sh | bash -s -- --auto

After install, ~/.local/bin/morie is a thin shim into the managed venv at ~/.venvs/morie. Full install instructions, channel comparison, and platform-specific notes are at rootcoder007.github.io/morie/#quick-start.

On minimal Linux containers (Alpine, slim Debian) that ship without curl, install it first: apt-get install -y curl or apk add curl. macOS already has curl built in.

Debian, Ubuntu, Raspberry Pi OS — plain pip in a venv

Debian-family systems mark the system Python "externally managed", so pip install morie stops with an error. A virtual environment is the supported way and needs nothing beyond the stock python3 (add sudo apt-get install -y python3-venv if venv is missing):

python3 -m venv ~/.venvs/morie
source ~/.venvs/morie/bin/activate
pip install -U morie
morie --version
morie interactive install

Activate the venv (source ~/.venvs/morie/bin/activate) in each new shell, or call ~/.venvs/morie/bin/morie directly.

Recommended — Windows

Windows doesn't ship curl, bash, python, or R, so the Linux/macOS one-liner above won't run there. The path that works on any Windows with no prerequisites:

1. Install Python from python.org/downloads — on the first installer screen, tick "Add python.exe to PATH" (skipping this is the No. 1 cause of python being "not recognized" in the terminal).
2. (Optional — for the R package) install R from cran.r-project.org/bin/windows/base.
3. Open PowerShell and install morie:

python -m pip install --upgrade pip
python -m pip install morie
python -c "import morie; print(morie.__version__)"
morie interactive install

For the R package, install rmorie (the R distribution of morie): Rscript -e "install.packages('rmorie', repos=c('https://rootcoder007.r-universe.dev','https://cloud.r-project.org'))"

Prefer a package manager? If winget --version works on your machine, winget install -e --id Python.Python.3.12 (and RProject.R) installs the prerequisites in one line each — but winget is absent from many Windows installs, so the installer steps above are the reliable default. The full Windows walkthrough, including fixes for common errors (python opening the Microsoft Store, PowerShell execution policy, long-path), is in INSTALLATION.md.

Python — Homebrew (macOS / Linuxbrew)

If you don't have Homebrew yet, install it first (macOS ships curl and bash, so this works out of the box):

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

Then:

brew tap rootcoder007/morie
brew install morie
morie interactive install

The tap repo is rootcoder007/homebrew-morie. It pulls morie's source distribution from PyPI and bundles a self-contained python@3.12 venv — no system Python required.

Python — PyPI (manual; requires pip already installed)

pip install morie
morie interactive install

Heads-up: Debian, Ubuntu and Raspberry Pi OS forbid pip outside a virtual environment (PEP 668). Use the venv three-liner above, or the one-liner installer, which sets one up for you. morie needs no NumPy or SciPy, so no compiled scientific stack has to be present.

Python — Docker (no local dependencies)

# Latest stable
docker run --rm ghcr.io/rootcoder007/morie:latest morie --help

# Pin to a specific version (recommended for reproducibility)
docker run --rm ghcr.io/rootcoder007/morie:1.4.0 morie --help

Published on every release with a versioned tag, a major.minor tag and :latest (linux/amd64). Requires only Docker — no Python, no pip.

The interactive layer — morie repl, morie exec, morie agent, morie tui

Every channel above leaves five modules out on purpose: they execute code that you or a model type, and package scanners flag that surface. Add them for your user in one step (the one-liner installer does it for you; a verb that needs them prints this command and, on a terminal, offers to run it):

morie interactive install

That fetches the files for your installed version from the tagged source on GitHub, checks each one against the SHA-256 manifest shipped inside the package, and stores them under ~/.local/share/morie/interactive (%LOCALAPPDATA%\morie\interactive on Windows). Nothing in site-packages changes. morie interactive status shows what is active and morie interactive remove takes it out again. The TUI also needs pip install "morie[interactive]". A source checkout has all of this already. Offline: morie interactive install --from path/to/morie/src/morie.

R package: rmorie — r-universe

The R distribution of morie is the rmorie package.

# rmorie comes from r-universe (prebuilt binaries for macOS and Windows,
# source on Linux); its companions rmoriebricklayer and rmoriedata are on CRAN
install.packages(
  "rmorie",
  repos = c(
    rootcoder007 = "https://rootcoder007.r-universe.dev",
    CRAN     = "https://cloud.r-project.org"
  )
)
install.packages(c("rmoriebricklayer", "rmoriedata"))

# The same R code also lives in this repository as r-package/morie (package
# name "morie"), tracking every commit here; build it from source with remotes
# (needs a C++ toolchain and rmoriebricklayer):
# install.packages("remotes")
remotes::install_github("rootcoder007/morie@v1.4.0", subdir = "r-package/morie")  # the tag of your morie

Or let morie run either install for you: morie r-install (r-universe) or morie r-install --github (this repository's R arm).

Quick start

import morie
from morie.data import load_dataset

# CPADS 2021-22 public-use microdata (40,931 rows): downloaded from
# open.canada.ca on first use (about a minute), then served from the
# local cache in ~/.cache/morie
df = load_dataset("ocp21")
print(df.shape)

# Welch's two-sample t-test, and the guide every morie.fn callable carries
from morie.fn import describe, welcht

result = welcht([5.1, 4.9, 5.6, 5.8, 6.0], [6.2, 6.8, 7.1, 6.5, 7.4])
print(result)
print(describe("welcht"))

From the terminal

Every feature is a verb of the morie command; morie --help lists them, morie cheatsheet fits them on one page, and the R package ships the same verbs as rmorie (see below).

morie list-modules                                   # the 23 analysis modules
morie run-module power-design --output-dir out/      # one module, its tables as CSV
morie explain power_two_proportion_gender.csv        # how to read an output table
morie list-datasets                                  # 71 catalog keys + the curated tables after login
morie pull ocp21 --out cpads.csv                     # the real CPADS PUMF, cached; modules use it from then on
morie pull --all --out datasets/                     # every catalog dataset
morie login --token                                  # the key issued at rmorie.com/access: hosted model tier + data.rmorie.com
morie login                                          # GitHub sign-in, for accounts that have one
morie login --email you@example.com                  # an emailed code instead: type it at the prompt
morie login --no-browser                             # server / SSH / no browser: prints a link + code for any device
morie pull chicago_crime/incidents --out incidents.csv   # a curated table (8.6M rows)
morie ask "which module fits a treatment-control design?"
morie provider set --base-url https://api.openai.com/v1 --key sk-...   # or your own model endpoint
morie emissions --seconds 5 --country CAN            # energy + CO2 of this machine, sealed in a capsule
morie pipeline --modules power-design -y             # modules + emissions tracking + capsule
morie verify-pollution --pollutant no2 --demo        # pollution -> health causal pipeline
morie selftest                                       # every subsystem, offline (no downloads)

Where the datasets come from

morie list-datasets prints a Route column next to every key. Of the 71 keys, 70 download themselves on first use and are cached: open.canada.ca (CPADS, CSADS, CSUS microdata and bootstrap weights), data.ontario.ca (the OTIS correctional tables and institution locations), Statistics Canada (CCHS), CIHI (the indicator library and its tables), Environment Canada's NAPS air-quality files, the Toronto Police ArcGIS feeds (the Canada-wide NAPS hourly keys are about 2 million rows: allow ten minutes and 1.5 GB), the Health Infobase tables from health-infobase.canada.ca with the data.rmorie.com copy as the fallback, the OTIS research environments from data.rmorie.com (R objects: rmorie loads them, morie saves them for R), and the reviewed SIU corpus from the rmoriedata package on CRAN (fetched as a source tarball, no R needed). One key, the MAPQ workbook, is your own file: put it under a data directory and point MORIE_DATA_DIR at it, keeping the relative path that morie list-datasets shows. The curated tables at data.rmorie.com join the list once a key (issued on request at https://rmorie.com/access) is stored with morie login --token.

from morie.data import list_rmoriedata, load_rmoriedata
[r["slug"] for r in list_rmoriedata()]        # 99 tables + 8 data dictionaries
siu = load_rmoriedata("siu_directors_reports")  # 5,157 reviewed SIU reports

Beyond the catalog, the project keeps 160 databases materialised from Google BigQuery public datasets, plus the Health Infobase tables and the OTIS research files (Chicago crime, EPA air quality, US census, FEC, FDA, NOAA, NHTSA, Hacker News, Ethereum, World Bank, ...), served from the edge at https://data.rmorie.com and opened by the key morie login --token stores (issued on request at https://rmorie.com/access, under https://rmorie.com/data-license). morie list-datasets shows their db/table keys with the route "data.rmorie.com"; morie pull db/table and load_dataset("db/table") fetch one (cached locally), and https://data.rmorie.com/browse runs SQL on any of them in the browser. They are rebuilt weekly without any project machine in the loop.

The R side has every verb

rmorie (r-universe) ships the same command line: rmorie::install_cli() puts rmorie on your PATH, and rmorie pull ocp21, rmorie run-module, rmorie emissions, rmorie verify-pollution, rmorie provider set, rmorie selftest and the rest behave like their morie twins. The two packages share the credentials file, the CPADS resolution order, the emissions CSV layout and the capsule format, so they mix in one workflow. morie r-install installs the R side from Python.

What's new

Per-release user-facing changes are now in WHATS_NEW.md. For the R-package changelog see r-package/morie/NEWS.md. For the planned roadmap see ROADMAP.md.

Documentation

Full documentation is at rootcoder007.github.io/morie.

• Website: https://rmorie.com — the MORIE family (rmorie, morie, rmoriebricklayer, rmoriedata) in one place.
• Curated data: https://data.rmorie.com — the BigQuery-built tables, opened by the same key, issued on request at https://rmorie.com/access; /browse for SQL in the browser.
• CLI reference: cli; learn pages for datasets, emissions and capsules and the R command line.
• Hosted LLM tier: https://llm.rmorie.com — the last resort behind morie ask, after a local Ollama and every key of your own. Keys are issued on request at https://rmorie.com/access and stored with morie login --token; morie login (GitHub) and morie login --email you@example.com still work for accounts that have them. The endpoints and the model list come from a signed services document the package verifies before use, so they can change without a release. On a server or over SSH, morie login --no-browser prints a link and a code to open on any device (1.4.0 opens no browser there on its own); morie models lists the models on your key and morie ask --model NAME picks one; see the hosted tier docs. The tier serves ollama.com cloud models and additional AI models (kimi-k2.6:cf, kimi-k2.7-code:cf, deepseek-v4-pro:cf, deepseek-v4-flash:cf, glm-5.2:cf, glm-5.3:cf, glm-5.3-flash:cf, gpt-oss-120b:cf, gpt-oss-20b:cf, llama-4-scout:cf, qwen3.8-27b:cf, nemotron-3-120b:cf and gemma-4-26b:cf); a rate-limited cloud model falls back to one of them.

Citation

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

Ruhela, V. S. (2026). morie: Multi-domain Open Research and Inferential Estimation. https://github.com/rootcoder007/morie

BibTeX:

@Manual{ruhela_morie_2026,
  title  = {morie: Multi-domain Open Research and Inferential Estimation},
  author = {Vansh Singh Ruhela},
  year   = {2026},
  url    = {https://github.com/rootcoder007/morie},
}

The single citation above covers both the R (r-package/morie/) and Python (src/morie/) implementations, which ship under the same version.

See CITATION.cff for machine-readable citation metadata (BibTeX, etc.) — that file is what GitHub's "Cite this repository" button consumes.

Acknowledgments

AI assistance

MORIE was developed with substantial assistance from frontier AI assistants. The author retains full responsibility for the code, the methods, and the scientific claims; AI assistance accelerated implementation but does not change the attribution of the work.

• Claude — Anthropic. Anthropic's Claude family (Opus, Sonnet, and Haiku across the 4.x and 5.x generations) was used extensively throughout development for code generation, refactoring, documentation, code review, and design discussions. Use was supported by Anthropic research-credit programs.

• Gemini and Vertex AI — Google. Google's Gemini 2.5 models (Pro and Flash) on the Vertex AI platform were used extensively for additional code generation, cross-checking Claude-generated code, multi-modal data analysis, and prototype evaluation. Use was supported by Google research-credit programs.

Funding and infrastructure

• Anthropic — Claude API research credits.
• Google — Gemini / Vertex AI research credits.

Data acknowledgments

Several MRM analyses use Statistics Canada and Health Canada Public Use Microdata Files (PUMFs) — including the Canadian Cannabis Survey (CCS), the Canadian Student Alcohol and Drugs Survey (CSADS), the Canadian Substance Use Survey (CSUS), the Canadian Alcohol and Drugs Survey (CADS), and the Canadian Postsecondary Education Alcohol and Drug Use Survey (CPADS) — along with Public Health Agency of Canada (PHAC) and Canadian Institute for Health Information (CIHI) aggregates. Although the analyses use Statistics Canada and Health Canada data, the analyses, interpretations, and conclusions are those of the author and do not represent the views of Statistics Canada or Health Canada. Ontario open data (OTIS, A01-RCDD release; via data.ontario.ca) and Toronto Police Service open data are used under the same standard disclaimer.

License

morie is licensed under the GNU Affero General Public License, version 3.0 or later (AGPL-3.0-or-later), on both the Python and R sides. The AGPL is a strong copyleft license: anyone who distributes a modified morie — or offers a modified morie to users over a network — must publish their source. Modifications and improvements cannot be kept secret or taken closed-source.

• Python and R packages (src/morie/, r-package/morie/) — AGPL-3.0-or-later. See LICENSE.
• Optional Linux kernel adjuncts (kernel-module/morie.c, daemon/morie_lsm.py) — GPL-2.0-only (the Linux kernel ABI requires GPL for loaded modules). These are NOT part of the R / Python distribution; they are separately-licensed, independently-distributed adjuncts. See kernel-module/LICENSE-GPL2.
• Papers, data and documentation — CC BY-NC-SA 4.0 (Creative Commons Attribution-NonCommercial-ShareAlike) unless explicitly marked otherwise.

Full detail in LICENSING.md.

Trust model — what the powerful features can do to your machine

morie is a developer/research toolkit with some intentionally powerful surfaces. They run local, user-supplied input by design — the only network-supplied code is the verified interactive layer below — but under an untrusted-input threat model each is high-impact, so every one defaults to the safe behaviour and gates the risky path behind a single, named, off-by-default environment knob:

morie doctor prints the current state of every knob so you can see the active trust posture at a glance. Only enable a knob for inputs you fully control.

Reporting issues / security

• General issues: GitHub Issues
• Security vulnerabilities: see SECURITY.md