Scaltair

Simple plotting library for Scala based on Vega-Lite, inspired by Python's Vega-Altair.

Plots can be rendered in the browser or from within a jupyter-notebook.

Example

Here is an example using Scaltair to quickly visualize and display a dataset with the native Vega-Lite renderer in the browser:

import scaltair.*
import scaltair.PlotTargetBrowser.given
import scaltair.vegalite.datasets.VegaDatasets

@main def examplePlot() =
  val data = VegaDatasets.loadCars().get
  val plot = Chart(data)
    .encode(
      Channel.X("Horsepower", FieldType.Quantitative),
      Channel.Y("Miles_per_Gallon", FieldType.Quantitative),
      Channel.Color("Origin", FieldType.Nominal)
    )
    .mark(Mark.Circle())
    .show()

Getting started

See the quickstart Guide.

Status of the project

The project is in an early stage and the api might still be subject to changes. It supports a subset of the functionality provided by vega-lite and altair. While it can already produce most plots used in a typical scientific context or data analysis task, it cannot create interactive plots nor does it support temporal data.

It should be easy to add the missing functionality if this is required. The project is designed in such a way, that it is easy to understand and extend. It does not make use of any fancy language features or libraries. Pull requests are very welcome.

Trying it out

While there is not yet an official version released, a snapshot release can be found on sonatype. To try it out, add the following dependency to your build.sbt:

resolvers +=  Resolver.sonatypeRepo("snapshots"), 
libraryDependencies += "ch.unibas.cs.gravis" %% "scaltair" % "0.2-SNAPSHOT"

If you want to use the jupyter-notebook integration, you also need to add the following dependency:

libraryDependencies += "ch.unibas.cs.gravis" %% "scaltair-jupyter" % "0.2-SNAPSHOT"

Adding new features to Scaltair

Scaltair consists of three main parts, which need to be extended to add new features:

1. The dsl for creating plots, which is defined in the scaltair package.
2. The converter, which converts the dsl to a vega-lite specification. It is defined in the file scaltair/DSLToVegaSpec.scala.
3. The json encoder, which converts the vega-lite specification to json. It is defined in the file scaltair/VegaLiteSpecJsonEncoder.scala.

To add a new feature, you need to extend all three parts. First, expose the new functionality in the DSL. Then add a new case to the converter, which converts the new dsl element to the vega lite specification. Finally, add a new case to the json encoder, to make sure that the field is correctly encoded to json.