⚠️ Important NoticeThis repository contains the original implementation of TINTO, an engine for converting tabular (tidy) data into synthetic images using dimensionality reduction and convolution techniques.
📦 We strongly recommend using the updated library TINTOlib, which includes:
- The original TINTO method
- Several additional methods such as IGTD, REFINED, BarGraph, DistanceMatrix, Combination, FeatureWrap, SuperTML, and BIE
- A much more user-friendly and flexible interface
- Complete and regularly updated documentation
- A free course with examples, notebooks, and video tutorials
🔄 TINTOlib is under active development and continues to receive improvements.
👉 For new projects and applications, we highly recommend switching to TINTOlib.
TINTO is a Python engine to transform Tidy Data (aka tabular data) into synthetic images, enabling CNN-based classification on non-visual datasets.
TINTO has a dedicated page in DeepWiki, where you can browse semantic documentation, use cases, FAQs, and more.
- 📊 Input: Tidy CSV files (target as last column)
- 🎯 Supports binary and multi-class classification
- 🔄 Dimensionality reduction: PCA and t-SNE
- 🖼️ Output: grayscale synthetic images with/without blurring
- ⚙️ Customizable image size and overlap strategies
- 🐍 Python 3.7+ compatible
TINTO-short-withSound.mp4
pip install -r requirements.txt
python tinto.py -h # view all available optionsExample - No Blurring:
python tinto.py iris.csv iris_imagesExample - With Blurring + t-SNE:
python tinto.py iris.csv iris_images_tSNE -B -alg t-SNE -oB maximum -px 30 -sB 5| sepal length | sepal width | petal length | petal width | target |
|---|---|---|---|---|
| 4.9 | 3.0 | 1.4 | 0.2 | 1 |
| 7.0 | 3.2 | 4.7 | 1.4 | 2 |
| 6.3 | 3.3 | 6.0 | 2.5 | 3 |
If you use TINTO, please cite:
@article{softwarex_TINTO,
title = {TINTO: Converting Tidy Data into Image for Classification with 2-Dimensional Convolutional Neural Networks},
journal = {SoftwareX},
author = {Manuel Castillo-Cara and Reewos Talla-Chumpitaz and Raúl García-Castro and Luis Orozco-Barbosa},
volume = {22},
pages = {101391},
year = {2023},
issn = {2352-7110},
doi = {https://doi.org/10.1016/j.softx.2023.101391}
}And for indoor localisation use-case:
@article{inffus_TINTO,
title = {A novel deep learning approach using blurring image techniques for Bluetooth-based indoor localisation},
journal = {Information Fusion},
author = {Reewos Talla-Chumpitaz and Manuel Castillo-Cara and Luis Orozco-Barbosa and Raúl García-Castro},
volume = {91},
pages = {173-186},
year = {2023},
doi = {https://doi.org/10.1016/j.inffus.2022.10.011}
}Learn how to load the generated images into CNNs:
TINTO is released under the Apache License 2.0.