Introduction

This documentation provides detailed instructions on how to install, configure, and use TINTOlib—a Python library designed to transform tabular data into synthetic images, enabling the application of advanced vision-based machine learning models on traditional data sets.

Installation

To use TINTOlib, first install it using PyPi:

(.venv) $ pip install TINTOlib

For more details, visit: TINTOlib on PyPI <https://pypi.org/project/TINTOlib>

Importing Methods

TINTOlib offers a variety of state-of-the-art methods for constructing synthetic images from tabular data. To use these methods, import the required models as shown below:

>>> from TINTOlib.tinto import TINTO
>>> from TINTOlib.igtd import IGTD

All the methods requires the same input format and the responses outputs will have the same format.

Available Methods

Models	Class	Hyperparameters
TINTO	`TINTO()`	problem, normalize, verbose, pixels, algorithm, blur, submatrix, amplification, distance, steps, option, times, train_m, zoom, random_seed
IGTD	`IGTD()`	problem, normalize, verbose, scale, fea_dist_method, image_dist_method, error, max_step, val_step, switch_t, min_gain, zoom, random_seed
REFINED	`REFINED()`	problem, normalize, verbose, hcIterations, n_processors, zoom, random_seed
BarGraph	`BarGraph()`	problem, normalize, verbose, pixel_width, gap, zoom
DistanceMatrix	`DistanceMatrix()`	problem, normalize, verbose, zoom
Combination	`Combination()`	problem, normalize, verbose, zoom
SuperTML	`SuperTML()`	problem, normalize, verbose, pixels, feature_importance, font_size, random_seed
FeatureWrap	`FeatureWrap()`	problem, normalize, verbose, size, bins, zoom
BIE	`BIE()`	problem, normalize, verbose, precision, zoom

Input Format

TINTOlib supports two primary data input formats:

Pandas DataFrame

Use a Pandas DataFrame with the target variable in the last column and features in other columns, all in numerical format.
Follow the Tidy Data principles.

CSV Files

Format your CSV with commas as separators and the target variable in the last column.
Ensure the first row contains feature names and all data is numerical.

Example Usage:

model = TINTO()
model.fit_transform('path_to_csv_file', 'result_folder_path')

Generating Synthetic Images

To generate synthetic images, TINTOlib provides fit, transform, and fit_transform methods:

Fitting the Model: The fit method trains the model on the tabular data:

model.fit(data)

Generating Synthetic Images: After fitting, use transform to generate and save images:

model.transform(data, folder)

Combining Fit and Transform: The fit_transform method combines training and image generation:

model.fit_transform(data, folder)

Parameters for Methods: - data: A path to a CSV file or a Pandas DataFrame containing the features and targets. - folder: Path where the synthetic images will be saved.

Output Format

The output of the model.transform(data,folder)() and model.fit_transform(data,folder)() method are synthetic images grouped in folders depending on their class value.

For example:

If the dataset is composed of 3 different classes, the method will create 3 folders and each folder will contain one synthetic image for each instance with that class.

—