nunchaku 0.8.0

To find the linear segment of a curve or dataset.

nunchaku: Dividing data into linear regions

nunchaku is a Python module for dividing data into linear regions. It answers two questions:

1. how many linear regions best fit the data without overfitting (by Bayesian model comparison);
2. given the number of linear regions, where the boundaries between them are (by finding the posterior of the boundaries).

Installation

For users, type in terminal

> pip install nunchaku

For developers, create a virtual environment and then type

> git clone https://git.ecdf.ed.ac.uk/s1856140/nunchaku.git
> cd nunchaku 
> poetry install --with dev 

Quickstart

Data x is a list or a 1D Numpy array, sorted ascendingly; the data y is a list or a Numpy array, with each row being one replicate of measurement.

>>> from nunchaku.nunchaku import nunchaku, get_example_data
>>> x, y = get_example_data()
>>> nc = nunchaku(x, y, prior=[-5,5]) # load data and set prior of slope
>>> # compare models with one, two or three linear regions
>>> num_regions, evidences = nc.get_number([1,2,3])
>>> # get the mean and standard deviation of the boundary points
>>> bds, bds_std = nc.get_iboundaries(num_regions)
>>> info_df = nc.get_info(bds)
>>> nc.plot(info_df)

Documentation

Detailed documentation is available on Readthedocs.

Citation

A preprint is coming soon.