https://github.com/anezovic1/i-encoding
https://pypi.org/project/iencoder/
https://github.com/anezovic1/i-encoding
Last synced: 27 days ago
JSON representation
https://pypi.org/project/iencoder/
- Host: GitHub
- URL: https://github.com/anezovic1/i-encoding
- Owner: anezovic1
- Created: 2024-05-31T15:59:06.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-12-10T21:42:52.000Z (7 months ago)
- Last Synced: 2025-01-16T02:22:01.726Z (6 months ago)
- Language: Jupyter Notebook
- Homepage: https://anezovic1.github.io/i-encoding/
- Size: 66.4 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# i-encoding
## Authors
- [Anida Nezovic](https://github.com/anezovic1)
- [Dr Aida Brankovic](https://scholar.google.it/citations?user=Lh3kj1MAAAAJ&hl=en)
- [Dr Jin Yoon](https://scholar.google.com.au/citations?user=Ol5i7bcAAAAJ&hl=en)
## Overview
The IEncoder is a custom encoder designed to transform categorical variables into numerical representations using a unique encoding technique. The fit_transform method is a key part of its functionality, combining the fitting and transformation processes into a single step.
### Purpose
The method encodes categorical features into numerical values by mapping each category to a unique angular representation. This approach ensures a compact and continuous numerical representation of categorical variables while excluding a target column.
### How It Works
The method starts by validating the input data X. It checks the format, dimensionality and ensures no invalid values (like NaN or inf) are present. Using the fit method, it identifies the categorical features in the dataset.
Each category is mapped to a unique angle in radians using a circular mapping strategy (2π divided by the number of categories).
If the target_column parameter is specified, the transformed dataset excludes the target column, as it is not meant to be encoded.
The final transformed dataset is returned as a pandas DataFrame, preserving the original feature names.
## Requirements
The package depends on the following libraries:
- numpy
- pandas
- scikit-learn
## Installation
i-encoder is on PyPi and can be installed using pip:
```bash
pip install iencoder
```
## Contact
If you have any questions, suggestions or feedback, feel free to reach out:
- Email: [email protected]
## License
This project is licensed under the [MIT License](LICENSE).