Quantile Dotplot

Implementation of a quantile dotplot in Python. Credit to Kay, Kola, Hullman, and Munson for providing papers, implementations, and pleasant talks on the subject.

These plots turn continuous distributions into discrete events, turning percentile calculations into counting. For example, with this chart, we can quickly tell that 10% of the time, the bus will be here in less than 8 minutes, and 30% of the time, in less than 10.

The above plot is made with the code

fig, ax = plt.subplots(figsize=(10, 4))
data = np.random.lognormal(mean=np.log(11.4), sigma=0.2, size=1_000_000)

ax = ntile_dotplot(data, dots=20, edgecolor="k", linewidth=2, ax=ax)

ax.set_xlabel("Minutes to bus")
for spine in ("left", "right", "top"):
    ax.spines[spine].set_visible(False)
ax.yaxis.set_visible(False)

where the ntile_dotplot command is provided by this library.

Installation

pip install git+https://github.com/ColCarroll/quantile_dotplot

Requires matplotlib for plotting and numpy for histogram work.

Quickstart

The library does its darndest to have osculating dots with reasonable aspect ratios, but errs on the side of osculating.

Here are three plots using the CASt dataset from Wolf et. al. (2004). This plots the brightness of 3,462 galaxies using an inverse logarithmic scale, so a galaxy with a value of 21 is 100 times brighter than one with a 26.

This first plot is made with

ntile_dotplot(data, dots=20, facecolor="gold", edgecolor="orangered", linewidth=4)

If we wish for more granularity, we can use

ntile_dotplot(data, dots=50, hist_bins=20, facecolor="gold", edgecolor="orangered", linewidth=4)

Once you are using lots and lots of dots, maybe turning integration into counting is a hard task, and you should think about using fewer dots (each dot here corresponds to 1% of the data!)

ntile_dotplot(data, dots=100, hist_bins=20, facecolor="gold", edgecolor="orangered", linewidth=4)

References:

Matthew Kay, Tara Kola, Jessica Hullman, Sean Munson. When (ish) is My Bus? User-centered Visualizations of Uncertainty in Everyday, Mobile Predictive Systems. CHI 2016. DOI: 10.1145/2858036.2858558.