Metadata-Version: 2.1
Name: emlens
Version: 0.0.11
Summary: A lightweight toolbox for understanding embedding space
Home-page: https://github.com/skojaku/emlens
Author: Sadamori Kojaku
License: MIT
Keywords: word embedding,graph embedding
Platform: UNKNOWN
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Science/Research
Classifier: Topic :: Software Development
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: faiss-cpu
Requires-Dist: tqdm
Requires-Dist: numba
Requires-Dist: numpy
Requires-Dist: scikit-learn
Requires-Dist: scipy

# emlens
[![Build Status](https://travis-ci.org/skojaku/emlens.svg?branch=main)](https://travis-ci.org/skojaku/emlens)
[![Unit Test & Deploy](https://github.com/skojaku/emlens/actions/workflows/main.yml/badge.svg)](https://github.com/skojaku/emlens/actions/workflows/main.yml)

A lightweight toolbox for analyzing embedding space

## Requirements
- Python 3.7 or 3.8 (may not work on 3.9)

## Doc

https://emlens.readthedocs.io/en/latest/

## Install

```bash
pip install emlens
```

`emlens` uses [faiss library](https://github.com/facebookresearch/faiss), which has two versions, `faiss-cpu` and `faiss-gpu`.
As the name stands, `faiss-gpu` can leverage GPUs, thureby faster if you have GPUs. `emlens` uses `faiss-cpu` by default to avoid unnecessary GPU-related troubles.
Yet, you can still leverage the GPUs (which is recommended if you have) by installing `faiss-gpu` by

*with conda*:
```bash
conda install -c conda-forge faiss-gpu
```

or *with pip*:
```
pip install faiss-gpu
```

### Installing emlens


## Maintenance

Code Linting:
```bash
conda install -y -c conda-forge pre-commit
pre-commit install
```

Docsctring: sphinx format

Test:
```bash
python -m unittest tests/simple_test.py
```


