OpenML.jl (Julia) Documentation¶

This is the reference documentation of OpenML.jl.

The OpenML platform provides an integration platform for carrying out and comparing machine learning solutions across a broad collection of public datasets and software platforms.

Summary of OpenML.jl functionality:

OpenML.list_tags(): for listing all dataset tags
OpenML.list_datasets(; tag=nothing, filter=nothing, output_format=...): for listing available datasets
OpenML.describe_dataset(id): to describe a particular dataset
OpenML.load(id; parser=:arff): to download a dataset

Installation¶

using Pkg
Pkg.add("OpenML")

If running the demonstration below:

Pkg.add("DataFrames") 
Pkg.add("ScientificTypes")

Sample usage¶

using OpenML # or using MLJ
using DataFrames

OpenML.list_tags()

Listing all datasets with the "OpenML100" tag which also have n instances and p features, where 100 < n < 1000 and 1 < p < 10:

ds = OpenML.list_datasets(
          tag = "OpenML100",
          filter = "number_instances/100..1000/number_features/1..10",
          output_format = DataFrame)

Describing and loading one of these datasets:

OpenML.describe_dataset(15)
table = OpenML.load(15)

Converting to a data frame:

df = DataFrame(table)

Inspecting it's schema:

using ScientificTypes
schema(table)

Public API¶

@docs OpenML.list_tags OpenML.list_datasets OpenML.describe_dataset OpenML.load