Introduction

Welcome to Superman! This library helps you use Manifold Learning algorithms in a functional, numpy-style manner. It is designed to be easy to use, well-documented, and fast.

Features

  • Clean, Pythonic API

  • Functional style

  • Abstracted dense/sparse operations

Installation

You can install the latest release from PyPI:

pip install superman

Or install from source:

git clone https://github.com/ovmurad/superman
cd superman
pip install -e.

Here’s a minimal example to get you started:

from src.geometry import Points
from src.data import load_swiss_roll

#returns tuple of points and intrinsic parameter
points_coords = load_swiss_roll(1000)
points = points_coords[0]
coords = points_coords[1]

#get the pairwise distance matrix with cityblock metric
dist = points.pairwise_distance(dist_type="cityblock")
print(dist)

Next Steps

  • Usage — learn how to use MyProject in more detail

  • src — explore the full API reference