vega-transform-omnisci-core

Vega transform for OmniSci Core.

Usage no npm install needed!

<script type="module">
  import vegaTransformOmnisciCore from 'https://cdn.skypack.dev/vega-transform-omnisci-core';
</script>

README

Vega Transform to Query OmniSciDB

npm version Test

Data transform to load data from an OmniSciDB database in Vega.

This package extends Vega's set of data transforms to support loading data from a database in Vega version 5.0 and higher.

Try this transform in our Observable demo or the scalable Vega demo app.

Usage Instructions

Install the transform with

yarn add vega-transform-omnici-core

To use the core transform, you must set the session. To create a session, create a connection and connect to it. Then assign the session to the Core transform with QueryCore.sesssion(session). the register the transform as querycore.

QueryCore.session(session);
transforms["querycore"] = QueryCore;

Here is a complete example.

import "@mapd/connector/dist/browser-connector";
import QueryCore from "vega-transform-omnisci-core";
import vega from "vega";

const connection = new window.MapdCon()
  .protocol("https")
  .host("metis.mapd.com")
  .port("443")
  .dbName("mapd")
  .user("mapd")
  .password("HyperInteractive");

// connect to core database and create a transform with a handle to the session
connection.connectAsync().then(session => {
  // pass the session to the core transform
  QueryCore.session(session);

  // register OmniSci Core transform
  vega.transforms["querycore"] = QueryCore;

  // now you can use the transform in a Vega spec
  const view = new vega.View(vega.parse(spec))
    .initialize(document.querySelector("#vis"));
  
  view.runAsync();
});

Vega Specifications

Once vega-transform-omnisci-core has been imported and registered, Vega specs can reference the transform and get data from OmniSciDB like this:

{
  "data": [
    {
      "name": "table",
      "transform": [{
        "type": "querycore",
        "query": "select count(*) from flights_donotmodify"
      }]
    }
  ]
}

Check out a complete demo at https://github.com/vega/scalable-vega.