Idris2-Jupyter-Vega

A Vega display plugin for the Idris 2 Jupyter kernel.

Wraps the ipyvega library.

Installation

Ensure that the Idris 2 Jupyter kernel, and the Idris 2 JSON Schema tool are installed, in a sibling directory of this repository.

Install the ipyvega library.

Note that the not needed in notebook >= 5.3 comment, in the ipyvega installation instructions, refers to the --sys-prefix flag, not the command as a whole.

Then run:

make install

Displaying graphs

Run the Idris 2 kernel with the idris2-jupyter-vega and contrib packages available.

Construct a VegaLite object:

import Language.JSON
import Idris2JupyterVega.VegaLite

export
barChart : String -> List (String, Double) -> VegaLite
barChart description vals = TopLevelSpec_0 $ MkTopLevelUnitSpec
    {Schema = Just "https://vega.github.io/schema/vega-lite/v5.json"}
    (Data_0 $ Data_0 $ DataSource_1 $ MkInlineData $ InlineDataset_3 $ map (\(name, x) => JObject [("a", JString name), ("b", JNumber x)]) vals)
    {description = Just description}
    {encoding = Just $ MkFacetedEncoding
        {x = Just $ PositionDef_0 $ MkPositionFieldDef
            {axis = Just $ Axis_0 $ MkAxis {labelAngle = Just $ LabelAngle_0 0}}
            {field = Just $ Field_0 "a"}
            {type = Just StandardTypeNominal}
        }
        {y = Just $ PositionDef_0 $ MkPositionFieldDef
            {field = Just $ Field_0 "b"}
            {type = Just StandardTypeQuantitative}
        }
    }
    (AnyMark_2 MarkBar)

Import the Idris2JupyterVega.VegaLite and Language.JSON modules, as well as any of your own definitions:

:module Idris2JupyterVega.VegaLite
:module Language.JSON

:let description : String
:let description = "A simple bar chart with embedded data."

:let vals : List (String, Double)
:let vals = [("A", 28), ("B", 55), ("C", 43), ("D", 91), ("E", 81), ("F", 53), ("G", 19), ("H", 87), ("I", 52)]

Execute the display function on your VegaLite object:

:exec display $ barChart description vals