Metadata-Version: 2.1
Name: data_science_pro
Version: 0.1.0
Summary: LLM-powered modular data science pipeline
Author: Your Name
Requires-Python: >=3.8
Description-Content-Type: text/markdown
Requires-Dist: pandas
Requires-Dist: scikit-learn
Requires-Dist: langchain
Requires-Dist: openai
Requires-Dist: python-dotenv
Requires-Dist: joblib
Requires-Dist: imbalanced-learn

# data_science_pro

A modular, LLM-powered data science pipeline for automated EDA, preprocessing, model selection, training, and testing.

## Installation

```bash
pip install .
```

## Usage

```bash
data-science-pro --help
```

Or run the pipeline interactively:

```bash
python pipeline.py
```

## Features
- Automated EDA and preprocessing
- LLM-powered suggestions and cyclic workflow
- Model selection, training, evaluation, and saving
- CLI entrypoint for easy usage

## Requirements
See `requirements.txt` for dependencies.

## Project Structure
- `api/` - LLM connector
- `cycle/` - Suggestion and control logic
- `data/` - Data loading, analysis, operations
- `modeling/` - Model training, evaluation, registry
- `utils/` - Utility files
- `pipeline.py` - Main pipeline class
- `test.py` - Example/test script
