Skip to main content

mosox

An LP matrix generator for GMPL (MathProg) and (by extension) AMPL.

Can be used in several ways:

  • Compile an MPS file for solving elsewhere
  • Solve directly with the HiGHS integration, no other binary needed
  • From Python, with data from pandas (or a .dat file), and scenario sweeps

It works for a subset of GMPL (specifically the subset required to run Osemosys). See Known limitations section below.

There are a number of examples of varying complexity in the examples/ directory.

Quickstart

Installation

Using Cargo

cargo install mosox

Using Homebrew (macOS)

brew tap carderne/mosox-tap
brew install mosox

Python

pip install mosox  # or: uv add mosox

Requires Python >= 3.11. Wheels are built for Linux x86_64, macOS arm64 and Windows x64; elsewhere pip builds from source, which needs a Rust toolchain.

Pre-built binaries (macOS, Windows, Linux)

Binaries are built for a small set of systems and architectures. Available to download (compressed) from the Releases page. Please choose the appropriate archive.

Usage overview

Usage overview:

mosox help

LP matrix generator for GMPL

Usage: mosox <COMMAND>

Commands:
  compile    Load and output to MPS
  solve      Solve with HiGHS
  normalize  Normalize an MPS file for diffing
  compare    Compare two normalized MPS files with epsilon tolerance
  help       Print this message or the help of the given subcommand(s)

Options:
  -h, --help     Print help
  -V, --version  Print version

Compile MPS

Compile MPS for one of the examples in this repo. Output will be written to the provided file:

mosox compile examples/basic/model.mod -o output.mps

Or an example with a separate data file, writing MPS to a file:

mosox compile examples/sets/model.mod examples/sets/data.dat -o output.mps

Solve a model

Results will be saved to the output file:

mosox solve examples/basic/model.mod -o output.txt

# Loading model from model.mod
# Generating matrix
# Matrix compiled in 666.625µs
# Matrix: 3 rows, 2 cols, 6 nonzero
# Solving matrix with HiGHS
# Solved in 1.939083ms
# Objective value: 2200
# Results output to output.txt

Python

Pass data as pandas DataFrames, on top of any data in the model file or a .dat file. Index columns are matched by set name (as in otoole), then VALUE.

import pandas as pd
import mosox

model = mosox.Model.from_file("examples/2d_params/model.mod")
cost = pd.DataFrame({"PLANTS": ["P1", "P2"], "WAREHOUSES": ["W1", "W1"], "VALUE": [9, 11]})

# `update` patches rows; dicts and scalars work too
solution = model.solve(mosox.Data().update({"cost": cost, "supply": {"P1": 120}}))
solution.status, solution.objective
solution["ship"]  # DataFrame: PLANTS, WAREHOUSES, VALUE, MARGINAL

Base data can come from a .dat file (Data.from_dat) or otoole (Data.from_otoole(*otoole.read(...))). Run a sweep over every combination of factors, in parallel:

model = mosox.Model.from_file("examples/osemosys_small/osemosys.mod")
base = mosox.Data.from_dat("examples/osemosys_small/ose_small.dat")
scenarios = mosox.product(
    carbon={"none": {}, "high": {"EmissionsPenalty": 100}},
    discount={r: {"DiscountRate": r} for r in [0.05, 0.10]},
)
run = model.run(scenarios, base, workers=4, keep=["NewCapacity"], store="results/")
for event in run:  # Started, Compiled, Finished, Failed, Skipped
    print(f"{event.done}/{event.total}", type(event).__name__, event.scenario.name)

results = run.results()  # or later: mosox.Results.open("results/")
results.summary          # a row per scenario: status, objective, timings
results["NewCapacity"]   # all scenarios, with scenario, carbon and discount columns

Development

Please install cargo-make:

cargo install cargo-make

The most useful dev commands are listed in Makefile.toml.

You can view available commands by running cargo make.

Formatting, linting

cargo make fmt
cargo make lint
cargo make ci  # fmt-check, lint, test (same as CI)

lint and test enable all features, including the Python bindings, which need Python >= 3.11. If your default python3 is older, point PyO3 at a newer one, e.g. export PYO3_PYTHON=$PWD/.venv/bin/python after uv sync.

Testing

In addition to other tests, this will run mosox against all the examples under examples/ and confirm that the output is identical to the existing MPS files.

cargo make test

This will additionally run a regression test against examples/osemosys_large if present.

cargo make testlarge

Python bindings

Requires uv. uv run builds the extension (unoptimised) as needed.

cargo make py  # ruff format, ruff check, ty check, pytest

Benchmarks

Run performance benchmarks across all examples (except osemosys_large):

cargo make bench

Include osemosys_large (requires the example to be present):

cargo make benchlarge

Run with glpsol comparison (requires glpsol on PATH):

cargo make benchglpsol

Performance vs glpsol (median, release build, Macbook Air M2)

Note that some of the performance advantage over glpsol may be caused by the limitations enumerated below.

Example rows/cols/nonzero N glpsol mosox Speedup
osemosys_small 6k/7k/15k 50 68ms 17ms 4.0x
osemosys_atlantis 180k/230k/510k 10 2.4s 373ms 6.5x
osemosys_large 1M/5M/12M 4 130s 20s 6.5x

Memory usage

The matrix generator uses very roughly 2,000 bytes per non-zero. This is a significant overhead over the 12 bytes per non-zero that would be needed in a format like CSC. However, as the table below shows, the matrix generator will rarely (if ever) use more memory than the solver.

Example Mosox matrix Glpsol matrix HiGHS solver Glpsol solver
osemosys_small 13 MB 12 MB 24 MB 16 MB
osemosys_atlantis 244 MB 287 MB 386 MB 416 MB
osemosys_large 5 GB 5.6 GB 6.5 GB ?

Known limitations

This list of limitations is made with reference to the GNU MathProg Language Reference which can be viewed here or downloaded from the original here.

It is intended that all of these will ultimately be supported, and most of them are "trivial" to add.

Functions

  • The following functions: abs, atan, ceil, cos, exp, floor, gmtime, length, log, prod, round, sin, sqrt, str2time, trunc, Irand224, Uniform, Normal.

Expressions

  • These arithmetic operators: less, div, mod
  • These symbolic operators: & (string concatenation)
  • These set expressions: conditional set expressions
  • String ordering comparisons (<, > etc. between symbolic values)

Sets

  • within and dimen are not enforced on computed sets that fail to evaluate at load (eg sets over tuple domains)

Parameters

  • Type specifiers, relational conditions and superset (in) expressions are only enforced on explicit data and constant defaults, not on computed params (:= expressions) or expression defaults
  • Plain-format data for multi-dimensional params without a slice (eg param p := a x 1 b y 2;) is not supported (it is caught as out-of-domain data)

Variables

  • Bounds specified as expressions (currently only constant accepted)

Constraints

  • Multiple expressions (comma-separated)

Other statements

Support for these statements is not planned (results should instead be parsed from the solver output): display, printf, for.

Metadata

Release files for mosox 0.8.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for mosox 0.8.2
File Size Uploaded Provenance Actions
mosox-0.8.2.tar.gz 556.3 kB With provenance Details for mosox-0.8.2.tar.gz

Built distributions (wheels)

Table of built distributions (wheels) for mosox 0.8.2
File Interpreter ABI Platform Actions
mosox-0.8.2-cp311-abi3-win_amd64.whl CPython 3.11 abi3 Windows x86-64 Details for mosox-0.8.2-cp311-abi3-win_amd64.whl
mosox-0.8.2-cp311-abi3-manylinux_2_39_x86_64.whl CPython 3.11 abi3 Linux glibc 2.39+ x86-64 Details for mosox-0.8.2-cp311-abi3-manylinux_2_39_x86_64.whl
mosox-0.8.2-cp311-abi3-macosx_11_0_arm64.whl CPython 3.11 abi3 macOS 11.0+ ARM64 Details for mosox-0.8.2-cp311-abi3-macosx_11_0_arm64.whl

Total release size: 39.2 MB

Release files / mosox-0.8.2.tar.gz

Download URL mosox-0.8.2.tar.gz
Size 556.3 kB
Tags Source
SHA-256 checksum
How to use checksums
b1fd7923b14a8a9115b36cf312843b963bbfce06514892b07cd9206fff1a6768
BLAKE2b-256 checksum
How to use checksums
2d03ced36bc45aaa5bad7dbf5b1b2ddc8fb3fec54ed6e3c9dfd7c4a47327d419
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 13, 2026.

Transparency log

Release files / mosox-0.8.2-cp311-abi3-win_amd64.whl

Download URL mosox-0.8.2-cp311-abi3-win_amd64.whl
Size 5.4 MB
Tags CPython 3.11 Windows x86-64 abi3
SHA-256 checksum
How to use checksums
8e281e11f45f83aeb1c04e5ffde78edcb86ddc82ea928faff0742bca87d236b5
BLAKE2b-256 checksum
How to use checksums
52eea1a9e371d952f0fbf19bc6e739a44dad72d0ef6b8a22d536625f9657a64a
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 13, 2026.

Transparency log

Release files / mosox-0.8.2-cp311-abi3-manylinux_2_39_x86_64.whl

Download URL mosox-0.8.2-cp311-abi3-manylinux_2_39_x86_64.whl
Size 28.2 MB
Tags CPython 3.11 Linux glibc 2.39+ x86-64 abi3
SHA-256 checksum
How to use checksums
6d484db3c5b9fc3ef0612fc95dbe5de3fd0cb1244fe4625130d774e1bce241b9
BLAKE2b-256 checksum
How to use checksums
e018b15f43d84076713475d222b8b00ac8bc30c064ef2d0fdf0ca4d4cdd9606b
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 13, 2026.

Transparency log

Release files / mosox-0.8.2-cp311-abi3-macosx_11_0_arm64.whl

Download URL mosox-0.8.2-cp311-abi3-macosx_11_0_arm64.whl
Size 5.0 MB
Tags CPython 3.11 abi3 macOS 11.0+ ARM64
SHA-256 checksum
How to use checksums
20b716c5d243b0ccd0b0a004aca53208039576f38105a91ffd6c7e030a40f634
BLAKE2b-256 checksum
How to use checksums
21f8bbdd8bb0d9b9317e031253f0ab38bf8416088b946da2cd96357f773cbfd4
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 13, 2026.

Transparency log

Security for mosox 0.8.2

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

All 4 files in this release have been uploaded with provenance metadata.

Select from 4 files

Attestation for mosox-0.8.2.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Nov 13, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page