API Reference /

datasets/

Datasets module for TuiML - Loading and generation of ML data.

The datasets module provides a unified interface for accessing classic machine learning datasets, loading data from various file formats, and generating synthetic data for benchmarking and experimentation.

Overview

This module is organized into three primary areas:

  1. Built-in Datasets: Easy access to standard datasets like Iris,
Diabetes, and CPU performance.
  1. Loaders: High-performance loaders for ARFF, CSV, Excel, Parquet,
and JSON formats.
  1. Generators: Synthetic data generation for classification,
regression, and clustering.

Loading a built-in dataset:

python 
>>> from tuiml.datasets import load_iris
>>> data = load_iris()
>>> X, y = data.X, data.y

Loading from a file with auto-detection:

python 
>>> from tuiml.datasets import load
>>> data = load('experiment_results.csv')

Generating synthetic data:

python 
>>> from tuiml.datasets import Blobs
>>> data = Blobs(n_samples=1000).generate()

Packages

builtin/
generators/
loaders/