multi.mjs

// This script exemplifies how to use the Range operator to create
// a multi-execution experiment.
//
// Make sure to have installed dependencies and have the required environment variables
// available, as described in the Quick start example:
//
// https://github.com/modelon-community/impact-client-js#quick-start
//
// Then run the example with: node multi.mjs

import dotenv from 'dotenv'
import {
    Analysis,
    Client,
    ExperimentDefinition,
    Model,
    Range,
} from '@modelon/impact-client-js'

// Load the .env file variables, install with: npm install dotenv
dotenv.config({ path: '../.env' })

const client = Client.fromImpactApiKey({
    impactApiKey: process.env.MODELON_IMPACT_CLIENT_API_KEY,
    serverAddress: process.env.MODELON_IMPACT_SERVER,
})

const WorkspaceName = 'test'
const workspace = await client.createWorkspace({
    name: WorkspaceName,
})

const analysis = Analysis.from({
    type: 'dynamic',
})
const model = Model.from({
    className: 'Modelica.Blocks.Examples.PID_Controller',
})

const experimentDefinition = ExperimentDefinition.from({
    analysis,
    model,
    modifiers: {
        variables: { 'PI.k': new Range(10, 100, 3).toString() },
    },
})

const experiment = await workspace.executeExperimentUntilDone({
    experimentDefinition,
    timeoutMs: 60 * 1000,
})

const cases = await experiment.getCases()

for (const simCase of cases) {
    const timeItems = await simCase.getTrajectories(['time'])
    const timeEndValue = timeItems[0].trajectory.pop()
    const inertiaItems = await simCase.getTrajectories(['inertia1.phi'])
    const inertiaEndValue = inertiaItems[0].trajectory.pop()
    const caseInput = await simCase.getInput()
    console.log(
        `End value for case with PI.k: ${caseInput.parametrization['PI.k']} at time ${timeEndValue}: ${inertiaEndValue}`
    )
}

await client.deleteWorkspace(WorkspaceName)