In our June Product Update, we shared that we made a strategic turn in our Python Integration
project. First, by making it more generic and thus enabling it to support scripting languages in general. Secondly, by decoupling the facilities from running these scripts from the AIMMS deployment options as those are widely available and needs vary greatly among customers making it impossible to keep ourselves.
As a result, we have now developed and released the following:
- A rebuild AIMMS HTTP Client library to interact with external (computational) services that support REST API calls
- A new AIMMS Data Exchange library for format conversions (e.g. JSON to AIMMS identifiers).
- A first set of ‘How-to’s that shows how you can integrate scripting languages, and specifically Python into AIMMS Apps (more ‘How-to’s will follow).
The combination of the libraries and documentation should give you ultimate flexibility in using the (scripting) service of the choice; now and in the future. It will allow you to leverage and customize your own scripting deployment environments and/or make use of specialized third party deployment services (such as Azure Functions, AWS Lambda, or any other services you might be aware of or already are using). Of course, the requirement will be that you need to have such a service in which you can host e.g. Python script, available to you.
In the “how-to’, we discuss how to add a scripting service to your model and how a setup could look like, both for development (local desktop) or deployment (on premise servers, or AIMMS cloud). Check it out by starting here: https://how-to.aimms.com/Articles/494/494-overview-aimms-ds-models.html