We have now released supported, documented versions of two libraries bridging the worlds of Python and AIMMS.
These will allow you to benefit from (a) the ubiquity of the Python language and its vast library collection and (b) from the proven AIMMS functionality for supporting industrial optimization models.
aimmspy
This library supports ‘Python in the lead’ scenarios. A Python application interacts directly with an in process AIMMS model for manipulation of the AIMMS data and running optimizations.
Primary use case for this is probably running AIMMS models ‘headless', for example as part of a larger application, in automated pipelines or for large-scale batch optimizations.
Key features:
- Interact with an AIMMS run-time from inside your Python script
- Read and write AIMMS identifier values
- Run procedures in AIMMS including solves
- Support for pandas and polars data frames
- Model declaration using AIMMS Developer
- Requires a valid AIMMS Developer license
Learn more from our documentation or reach out to support@aimms.com.
pyaimms
This AIMMS library supports ‘AIMMS in the lead’ scenarios, calling Python from within an AIMMS model.
Primary use case for this is probably outsourcing specialized calculations to Python. Think of Machine Learning, demand forecasting, data cleansing, data wrangling, etc.
Key features:
- Execute Python scripts or single statements
- Read and write AIMMS identifier values in Python
- Support for pandas and polars data frames
- Specify the Python version and all dependencies required using
- Can run on AIMMS Cloud, using a pyproject.toml file to specify the Python version and all dependencies
Learn more from our documentation or reach out to support@aimms.com.