GSEGUtils 0.5.2

General utility functions and classes for GSEG research/projects

GSEGUtils

GSEGUtils provides some tools and functionality that are used across other development or research projects. These are particularly focussed in the direction of research within the group.

The main goal if this repo is to help software development efforts and allow more focus on algorithm development, data processing and research rather than code debugging and maintenance.

Currently this module includes:

Package	Description
BaseArrays	Sub-classable array objects with automated validation on type, shape and attributes
Lazy Disk Cache	Automated data offloading for memory management
Util	Includes utility functions such as angle unit conversions
Validators	Helper functions for performing validation and normalisation

Installation

This can be done directly from the repository directory:

pip install git+ssh://[e-mail]/gseg-ethz/GSEGUtils.git

or by cloning it

git clone https://github.com/gseg-ethz/GSEGUtils.git
cd GSEGUtils
pip intstall .

Usage

Accessing the different classes can then be easily done by importing from GSEGUtils:

>>> from GSEGUtils.base_arrays import BaseArray, ArrayNx3, BaseVector
>>> import numpy as np
>>> data = np.random.rand(10,20,3)
>>> array = BaseArray(data)

>>> array.shape
(10, 20, 3)

>>> np.all(array == data)
np.True_

>>> a = ArrayNx3(np.random.rand(10, 3))
>>> b = a + 10
>>> np.all(b == a + 10)
np.True_

coords3D = Array_Nx3(np.random.rand(20, 3))
coords3d.arr = np.random.rand(100)
# RAISES ERROR -> invalid shape not of (N, 3), expected (N,)

coords3d.arr = np.random.rand(100, 3)
# Valid as the array is of shape Nx3

Publication Policy

This package is internal to ETH Zurich's GSEG group and is not published to PyPI. The release-please workflow produces build artefacts attached to GitHub Releases; internal users install via pip install <wheel-url> from those releases. There is no PyPI publish step; introducing one requires explicit ADR-level review.