ridgeplot-py 0.6.4

Plotting ridgeplots with matplotlib

ridgeplot-py

This is a simple module for plotting ridgeplot with the scipy ecosystem.

Ridgeplot is a great data visualization technique to compare distributions from multiple groups at the same time, and was first introduced in 2017 as joy plot:

I hereby propose that we call these "joy plots" #rstats https://t.co/uuLGpQLAwY

— Jenny Bryan (@JennyBryan) April 25, 2017

ridgeplot-py provides a simple API to produce matplotlib-compatible ridgeplots, as well as a handy ColorEncoder class with scikit-learn syntax for manipulating color annotations in a consistent way [through out manuscripts or presentations].

Install

git clone [e-mail]:wckdouglas/ridgeplot-py.git
cd ridgeplot-py
python setup.py install 

or via conda:

conda install -c conda-forge ridgeplot-py

or via pypi:

pip install ridgeplot-py

Usage

from ridgeplot import ridgeplot
from ridgeplot.colors import ColorEncoder, ColorPalette
import numpy as np
import matplotlib.pyplot as plt

# mocking some data
# the input data should be a dict of
# - keys: group names for the distributions
# - values: list of values 
data = {}
for i in range(8):
    data['data_{}'.format(i)] = np.random.randn(100) * (i+1)

# make the plot
fig = plt.figure()
ax = fig.add_subplot(111)
ridgeplot(
    ax, 
    data, 
    xlim=(-20,20), 
    label_size=15
)

Example

A notebook showing quick howto is included in this repo!

Build on Apple silicon

scipy may cause error and may be able to solved by the this stackoverflow answer:

brew install openblas
export OPENBLAS="$(brew --prefix openblas)" 
poetry install