Welcome to LinkingLines!

Read the Full documentation on ReadtheDocs!

1. Introduction

Welcome to the documentation for our Python module that performs the Hough Transform on lines from geospatial data, clusters it using Agglomerative Clustering. This module also includes custom plotting scripts and feature extraction methods to help you analyze and visualize your data effectively.

This code was used to create the results published in Kubo Hutchison et al., 2023. Initially, it was designed to link together mapped dike segments in Cartesian space to find their true lengths. This code can be applied to any linear features including roads, fractures, and other types of linear data.

Data Clustering: Apply Agglomerative Clustering to group similar data points, this can be used for data reduction, analysis, and mapping.
Data Visualization: Custom plotting scripts help you visualize and analyze your data, making it easier to identify patterns and anomalies.
Feature Extraction: Extract meaningful features from clustered data to perform further analysis, such as linear or radial type features.

Full documentation can be found on ReadTheDocs

2. Installation

To use this module, make sure you have Python installed (preferably Python 3.x). You can install the required packages using pip:

pip install linkinglines

3. Quick Start

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import linkinglines as ll

data=ll.readFile('path/to/data')

dtheta=2 #degrees
drho=500 #meters

dikeset, Z=ll.AggCluster(data)
lines,evaluation=examineCluster(data)
fig,ax=DotsLines(lines, ColorBy='AvgTheta')

We have three examples:

In-depth tutorial with Hough transform, clustering, and feature extraction using Spanish Peaks Data CSV file.
Hough Transform and feature extraction on Venus lineament data shape file.
Hough Transform on fracture data geoJSON.

Data from:

4. Contributing Guidelines

If you find bugs or issues please open up an issues. We also welcome requests for additional features.

If you would like to contribute code please do so in a separate branch and open up an issue describing your contribution.

git clone git@github.com:USER/LinkingLines.git
git checkout my-development-branch

We recommend using a virtual environment to manage packages. We use poetry to manage dependencies and building. See more about poetry.

pipx install poetry # you may need to install pipx first

cd linkinglines # go to the repo

poetry install --with test,dev # install in editable mode

# add your code

poetry run pytest  # test code locally

Before submitting your pull request please verify the following:

Code is documented in NumPy Docstring Style
Code is tested and passes test
- To run the tests please go to "/tests" and run poetry run pytest or pytest
- Add your test code to any file with the name test
- More here on pytest and testing practices
Open an issue and pull request
After your pull request the code will be reviewed by maintainer (namely @aikubo).
After passing review and automated tests it will be added to the next release and published to pypi.

hugoledoux / LinkingLines