Metadata-Version: 2.4
Name: numpy-dtype-utils
Version: 0.3.4
Summary: Advanced data type handling, gaze estimation, and dataset utilities for NumPy.
Home-page: https://github.com/yourusername/numpy-dtype-utils
Author: Takumi Sekiguchi
Author-email: [EMAIL_ADDRESS]
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Intended Audience :: Science/Research
Classifier: Topic :: Scientific/Engineering
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy
Requires-Dist: pandas
Requires-Dist: scipy
Requires-Dist: matplotlib
Requires-Dist: seaborn
Requires-Dist: scikit-learn
Requires-Dist: torch
Requires-Dist: fastapi
Requires-Dist: uvicorn
Requires-Dist: Pillow
Requires-Dist: pydantic
Requires-Dist: streamlit
Provides-Extra: test
Requires-Dist: pytest; extra == "test"
Dynamic: author
Dynamic: author-email
Dynamic: classifier
Dynamic: description
Dynamic: description-content-type
Dynamic: home-page
Dynamic: license-file
Dynamic: provides-extra
Dynamic: requires-dist
Dynamic: requires-python
Dynamic: summary

# numpy-dtype-utils

Advanced data type handling and dataset utilities for NumPy.

## Features

- **Efficient Data Loading**: Optimized loading for structured datasets.
- **DType Inspection**: utilities to analyze numpy dtypes (inspect.py).
- **Safe Type Casting**: Robust type conversion and promotion helpers (cast.py).
- **Data Cleaning Pipelines**: Streamlined handling of missing values, outliers, and duplicates.
- **Visualization Tools**: Integrated plotting for correlation matrices and distributions.
- **Clustering Support**: K-means clustering for spatial data analysis.

## Installation

```bash
pip install numpy-dtype-utils
```

## Usage

```python
from numpy_dtype_utils import Dataset, get_dtype_info, safe_cast
import numpy as np

# --- Dataset Utility ---
# Initialize dataset loader
dataset = Dataset("/path/to/data")

# Load and clean data
dataset.load(max_days=5).clean(remove_outliers=True)

# Visualize distributions
dataset.plot_distributions()

# --- Type Inspection ---
dt_info = get_dtype_info('float32')
print(dt_info)  # {'name': 'float32', 'itemsize': 4, ...}

# --- Safe Casting ---
arr = np.array([1, 2, 3], dtype='int32')
casted = safe_cast(arr, 'float64')
```

## License

MIT
