index.md

@sdeverywhere/runtime

High-level API

Create a ModelRunner around a GeneratedModel to run that model either synchronously or asynchronously.

• ModelRunner
• createSynchronousModelRunner (see @sdeverywhere/runtime-async package for asynchronous variant)

Model Scheduler

Optional: Create a ModelScheduler or MultiContextModelScheduler around a ModelRunner to simplify the process of running the model on demand. These classes are especially useful in web applications built with a reactive framework.

NOTE: The main difference between these two classes is that MultiContextModelScheduler allows for using multiple "contexts" (i.e., distinct sets of inputs and outputs) with the same underlying model instance, while ModelScheduler effectively only supports a single context. These two APIs will be combined prior to version 1.0 of the runtime package.

• ModelScheduler
• MultiContextModelScheduler

Inputs

The InputValue API is generally only needed when working with the ModelScheduler or MultiContextModelScheduler classes.

• InputValue
• InputCallbacks
• createInputValue

Outputs

Use the Outputs class to access the model outputs that are stored after each model run by the ModelRunner.

• Outputs
• Series
• Point

---

Low-level API

Note: Most developers will not need to interact with these low-level APIs directly. These are exposed only to show the shape of models that are generated by the SDEverywhere compiler.

API for generated JavaScript models

• JsModel

API for generated WebAssembly models

• WasmModule