Compute-to-Data with Nautilus

Compute-to-Data (CtD) allows you to generate value from data while ensuring it remains private. In the Pontus-X ecosystem, Nautilus provides a compute function to start and manage compute jobs. This chapter guides you through starting a new compute job, monitoring its status, and retrieving the results.

Introduction to Compute-to-Data

Compute-to-Data enables the analysis of data without exposing it. This functionality is crucial for maintaining data privacy and security while still extracting valuable insights.

Starting a New Compute Job

Starting a compute job with Nautilus is straightforward. The basic configuration requires only the identifiers of both the dataset and the algorithm.

1. Setup and Initialization: Begin by setting up your environment and creating a Nautilus instance.

   import { Nautilus } from '@deltadao/nautilus'
   import { providers, Wallet } from 'ethers'
    
   const provider = new providers.JsonRpcProvider('https://rpc.dev.pontux-x.eu')
   const signer = new Wallet('0x...', provider)
    
   // Create the Nautilus instance
   const nautilus = await Nautilus.create(signer)

2. Define Dataset and Algorithm: Specify the DIDs (Decentralized Identifiers) for both the dataset and the algorithm.

   const dataset = {
       did: 'did:op:123abc...'
   }
   const algorithm = {
       did: 'did:op:456def...'
   }

3. Start the Compute Job: Use the nautilus.compute() function to start the compute job.

   const computeJob = await nautilus.compute({
       dataset,
       algorithm
   })

4. Store the Job ID: Save the job ID for future reference.

   const { jobId } = computeJob

Monitoring Compute Job Status

Once you have a reference to a compute job, you can monitor its status using the getComputeStatus() function.

1. Get Compute Job Status: Retrieve the status of the compute job using its job ID and the provider URI.

   const computeJobStatus = await nautilus.getComputeStatus({
     jobId, // Using our extracted jobId
     providerUri: 'https://v4.provider.oceanprotocol.com/'
   })

Retrieving Compute Job Results

After a compute job has finished, you can access the results using the getComputeResult function.

1. Get Compute Job Results: Obtain the URL to retrieve the compute job results.

   const computeResultUrl = await nautilus.getComputeResult({
     jobId, // Use your previously saved jobId
     providerUri // Use the provider as described above
   })

2. Fetch the Data: Use the generated one-time URL to fetch the data.

   const data = computeResultUrl && await fetch(computeResultUrl)

Example Code

Here's a complete example demonstrating the compute-to-data process using Nautilus:

import { Nautilus } from '@deltadao/nautilus'
import { providers, Wallet } from 'ethers'
 
const provider = new providers.JsonRpcProvider('https://rpc.dev.pontux-x.eu')
const signer = new Wallet('0x...', provider)
 
// Create the Nautilus instance
const nautilus = await Nautilus.create(signer)
 
// Define dataset and algorithm
const dataset = {
    did: 'did:op:123abc...'
}
const algorithm = {
    did: 'did:op:456def...'
}
 
// Start the compute job
const computeJob = await nautilus.compute({
    dataset,
    algorithm
})
 
// Store the job ID
const { jobId } = computeJob
 
// Get the status of the compute job
const computeJobStatus = await nautilus.getComputeStatus({
  jobId,
  providerUri: 'https://v4.provider.oceanprotocol.com/'
})
 
// Get the compute job results
const computeResultUrl = await nautilus.getComputeResult({
  jobId,
  providerUri
})
 
// Fetch the data
const data = computeResultUrl && await fetch(computeResultUrl)

Deep Dive

For a detailed look into the compute APIs and their capabilities, refer to the following resources:

• nautilus.compute() - Full API Documentation
• nautilus.getComputeStatus() - Full API Documentation
• nautilus.getComputeResult() - Full API Documentation
• Compute Examples - Code Examples

Conclusion

Compute-to-Data in the Pontus-X ecosystem enables secure and private data analysis, maintaining data privacy while extracting valuable insights. By following this guide, you can efficiently start, monitor, and retrieve compute jobs using the Nautilus interface, leveraging the full potential of the Pontus-X platform.