Skip to main content

Knee or elbow detection for curves

Project description

kneebow

Find the knee of a curve or the elbow of a curve.

Codacy Badge Build Status codecov

How it works

kneebow builds upon a very simple idea: if we want to find the elbow of a curve, we can simply rotate the data so that curve looks down and then take the minimum value. If we want to find the knee of the curve, we take the maximum value instead. It's as simple as that.

For more info, see this answer on the Data Science StackExchange.

Installation

You can install the package via pip:

pip install kneebow

Alternatively, you can also install the latest version from GitHub:

pip install git+https://github.com/georg-un/kneebow.git

Usage

Let's assume, we try to find the elbow of the following data:

import numpy as np
data = np.array([[1, 1], [2, 2], [3, 3], [4, 4], [5, 5], [6, 6], [7, 7], [8, 8],  # linear until (8,8)
                 [9, 16], [10, 32], [11, 64], [12, 128], [13, 256], [14, 512]])   # exponential afterwards

Let's have a peak how this data looks like:

data_plot

To find the elbow, we create an instance of the Rotor class and use its fit_rotate method:

from kneebow.rotor import Rotor
 
rotor = Rotor()
rotor.fit_rotate(data)

Now we can get the index of the elbow as follows:

elbow_idx = rotor.get_elbow_index()
print(elbow_idx)  # 11

The Rotor class also comes with plot methods to inspect the data visually together with the estimated elbow/knee:

rotor.plot_elbow()

rotor_plot

Citation

If you need to cite this package, you can do so as follows:

BibTeX

@misc{kneebow,
  title={ {kneebow}: Knee or elbow detection for curves},
  author={Unterholzner Georg},
  year={2019},
  howpublished={\url{https://github.com/georg-un/kneebow}},
}

Note: Make sure to import the url package with: \usepackage{url}.

APA/Harvard

Unterholzner Georg. (2019). kneebow: Knee or elbow detection for curves. https://github.com/georg-un/kneebow.

License

Distributed under the MIT License. See LICENSE for more information.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

kneebow-1.0.1.tar.gz (4.5 kB view hashes)

Uploaded Source

Built Distribution

kneebow-1.0.1-py3-none-any.whl (4.7 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page