shadyquant😎

This python package allows you to quantile and plot lines where you have multiple samples, typically for visualizing uncertainty. Your data should be shaped (N, T), where N is the number of samples, T is the dimension of your lines.

Install

pip install shadyquant

Example

Consider you have 100 lines that you want to compute confidence intervals (quantiles) on:

import matplotlib.pyplot as plt
import numpy as np

x = np.linspace(0,2,1000)
y = np.sin(x)
plt.plot(x,y)

w = np.random.normal(size=100)**2
traj = y + y * w[:, None]
plt.plot(traj.T, color='C0')
plt.show()

You can use shadyquant to plot a 95% confidence interval:

import sq
sq.traj_quantile(traj)

You can also do weighted quantiling if you have weights attached to each line

sq.traj_quantile(traj, weights=w)

You may want to do fancy shading, which just plots a series of quantiles as polygons with transparency. The quantils overlap, which gives a nice blending. The outer edges of the polygons still correspond to the 95% confidence interval

sq.traj_quantile(traj, fancy_shading=True)

Here are some further options you can modify:

plt.figure(figsize=(8, 3))
ax = plt.gca()
sq.traj_quantile(
    traj,
    ax=ax,
    fancy_shading=True,
    lower_q_bound=1 / 3,
    upper_q_bound=2 / 3,
    levels=100,
    color="red",
    alpha=0.01,
)