substratumx 0.1.6

An NdArray library supporting numerical computation and spatial analysis.

Substratum

Substratum is a rust package designed for the dubious python library. Dubious was a project with the personal constraint that no libraries can be used other than numpy. Substratum is my attempt to rewrite some of numpy's features to replace numpy as dubious's lone dependency. I also intend to expand Substratum to include more features, although it is not intended to be a full fledged numpy replacement.

The core rational behind this library was to get a better grasp of how libraries I use on a regular basis work, and learn how to write python bindings to offload computationally expensive tasks to tools better suited. I chose rust because I have read the book around a year prior to this project, and pyo3 bindings are relitively easy to get working.

Status

This is largely a learning project and the API is subject to change. We achieve similar performance to numpy (beating them in a rare few) across most operations but basic broadcasting arithmetic is around 4x slower in most cases. I intend to add a few unsafe methods to speed things up where applicable but I don't intend to optimize much further at this stage.

Installation

pip install substratumx

You can also build with maturin build --release assuming maturin is installed.

Quickstart

import substratum as sx

a = sx.Array([2, 2], [1.0, 2.0, 3.0, 4.0])
b = sx.Array([2, 2], [5.0, 6.0, 7.0, 8.0])

print(f"a @ b = {(a @ b).tolist()}")

Output: a @ b = [[19.0, 22.0], [43.0, 50.0]]

Features

• Array, an N-dimensional array object
• Array broadcasting and trig methods
• matrix operations and constructors
• cholesky, qr and eigen decomposition
• KDTree, BallTree & VPTree with knn radius and KDE queries
• Statistical methods (mean, median, var, quantile)
• Pearson and Spearman correlation

Top-level

• substratum.Array

Modules

• substratum.linalg
• substratum.stats
• substratum.random
• substratum.spatial

Examples

import substratum as sx

a = sx.Array([2, 2], [1.0, 2.0, 3.0, 4.0])
b = sx.Array([2, 2], [5.0, 6.0, 7.0, 8.0])

print(f"a + b = {(a + b).tolist()}")
print(f"a * b = {(a * b).tolist()}")

Output: a + b = [[6.0, 8.0], [10.0, 12.0]] a * b = [[5.0, 12.0], [21.0, 32.0]]

import substratum as sx

gen = sx.Generator.from_seed(123)

uniform = gen.uniform(0.0, 1.0, [2, 3])
print(f"Uniform [0, 1): {uniform.tolist()}")

normal = gen.standard_normal([2, 3])
print(f"Standard normal: {normal.tolist()}")

Output: Uniform [0, 1): [0.19669435215621578, 0.9695722925002218, 0.46744032361670884, 0.12698379756585432] Standard normal: [-0.0008585765206425146, 1.4733334715623352, -1.16180050645278, -0.772101732825336]

Modules

Random

Generator object that can sample from uniform, normal, lognormal, gamma and beta distributions. Support for additional distributions is planned.

Linalg

• Standard matrix methods and constructors.
• cholesky and eigen and qr decomposition.

Stats

• Basic statistical methods for Array objects, mean, var and quantile.
• Pearson and Spearman correlation.

Spatial

• KDTree kNN, Kernel Density Estimation and radius queries.
• BallTree with kNN, Kernel Density Estimation and radius queries.
• VPTree with kNN, Kernel Density Estimation and radius queries.