GAMSPy: Mathematical optimization in Python
GAMSPy is an open-source Python package for building and solving mathematical optimization models. Its algebraic, indexed API supports linear, mixed-integer, nonlinear, quadratically constrained, and complementarity models while fitting naturally into Python data and analytics workflows.
Installation
pip install gamspy
GAMSPy requires Python 3.10 or newer. Installation includes the local execution runtime and a set of default solvers, so you do not need to install GAMS separately to run the example below. The included demo license is sufficient for this small model.
Your first model
This production-mix model chooses how many desks and chairs to make to maximize profit within 40 available labor hours:
import gamspy as gp
m = gp.Container()
products = gp.Set(m, records=["desk", "chair"])
profit = gp.Parameter(
m,
domain=products,
records=[("desk", 100), ("chair", 40)],
description="profit per unit",
)
hours = gp.Parameter(
m,
domain=products,
records=[("desk", 4), ("chair", 1)],
description="labor hours per unit",
)
production = gp.Variable(
m,
domain=products,
type="Positive",
description="units to produce",
)
capacity = gp.Equation(m)
capacity[...] = gp.Sum(products, hours[products] * production[products]) <= 40
model = gp.Model(
m,
equations=[capacity],
problem=gp.Problem.LP,
sense=gp.Sense.MAX,
objective=gp.Sum(products, profit[products] * production[products]),
)
model.solve()
print(f"Maximum profit: {model.objective_value:.0f}")
Maximum profit: 1600
The same workflow—define indexed data, variables, equations, and an objective, then solve—scales to larger and more complex models.
Why GAMSPy
- Algebraic, indexed modeling: Express whole families of variables and constraints over sets instead of constructing each scalar expression with Python loops. Model definitions stay close to their mathematical notation.
- Multiple problem classes: Build linear, mixed-integer, nonlinear, quadratically constrained, complementarity, and related model types with a consistent API.
- Choice of solver: Use compatible installed solvers without rewriting the model. GAMSPy provides access to both open-source and commercial solvers, subject to their license terms.
- Python workflows: Load model data from Python lists, NumPy arrays, or pandas objects, and work with results as pandas DataFrames.
Licensing and free use
The modeling package and the components used to execute and solve a model have different licensing considerations:
- GAMSPy source code: The Python source in this repository is available under the MIT License.
- Demo license: A time- and size-limited demo license is included with GAMSPy and can solve small models such as the example above. See the demo license limitations for current details.
- Free personal license: Individuals can generate a free license for non-commercial learning, experimentation, hobby projects, and prototypes. It includes a selection of free and open-source solvers; see the free-license overview for the applicable terms.
- Free academic GAMSPy license: Eligible students, teachers, researchers, and academic staff can obtain a free license for academic, non-commercial, non-production use. It includes selected commercial solvers without model-size limits. See the current included-solvers overview.
- Commercial license: Commercial and production use requires an appropriate license. Contact sales@gams.com or use the GAMS contact form and select Licensing.
Sign up and generate a free personal or academic license at the GAMS Portal, then follow the GAMSPy license installation instructions.
Solver availability depends on both the installed solver package and the active license. Runtime and solver components may have terms separate from the MIT-licensed Python source.
Relationship to GAMS
GAMSPy builds on GAMS model-generation and execution technology. You write and manage the model in Python; GAMSPy translates its algebraic representation for the execution system and dispatches it to a compatible solver. No knowledge of the GAMS language is required, and the standard pip install gamspy installation provides the local runtime needed to get started.
GAMS and GAMSPy licenses are separate. An existing GAMS license is not automatically a GAMSPy license; see the installation and licensing guide for the available options and exceptions.
Learn more
- Documentation
- Quick start guide
- Open the quick start in Colab
- Example model library
- API reference
- Licensing information
Community and contributing
For usage questions, visit the GAMSPy documentation or the GAMS Forum. Report bugs and request features through the GitHub issue tracker. Contributions are welcome; see the contributing guide to get started.
1 maintainer
Metadata
1 maintainer
Release files for gamspy 1.28.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| gamspy-1.28.0.tar.gz | 329.6 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| gamspy-1.28.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 721.1 kB
Release files / gamspy-1.28.0.tar.gz
| Download URL | gamspy-1.28.0.tar.gz |
|---|---|
| Size | 329.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
b12c48341ce00001564d32cab25aa7f183532dc619867fe6d1e4eeda4595651e
|
|
BLAKE2b-256 checksum How to use checksums |
0911d61cee955ab5ce8842c547f68b5e7abb3febfec9337b5080898c23166f58
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.12.15 {"installer":{"name":"uv","version":"0.12.15","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Debian GNU/Linux","version":"13","id":"trixie","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Release files / gamspy-1.28.0-py3-none-any.whl
| Download URL | gamspy-1.28.0-py3-none-any.whl |
|---|---|
| Size | 391.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
d4733423854e04f6d9cbb1e83a8c80292eb749b7497f8570ac8c830c9565fcbe
|
|
BLAKE2b-256 checksum How to use checksums |
b0f902ef168aac9bb0072c138bcc7be0c117f7b2ad3441960d018f565d698862
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
uv/0.12.15 {"installer":{"name":"uv","version":"0.12.15","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Debian GNU/Linux","version":"13","id":"trixie","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
