plotnine.geom_boxplot

geom_boxplot(mapping=None, data=None, **kwargs)

Box and whiskers plot

{usage}

Parameters

width : float = None: Box width. If None, the width is set to 90% of the resolution of the data. Note that if the stat has a width parameter, that takes precedence over this one.
outlier_alpha : float = 1: Transparency of the outlier points.
outlier_color : str | tuple = None: Color of the outlier points.
outlier_shape : str = "o": Shape of the outlier points. An empty string hides the outliers.
outlier_size : float = 1.5: Size of the outlier points.
outlier_stroke : float = 0.5: Stroke-size of the outlier points.
notch : bool = False: Whether the boxes should have a notch.
varwidth : bool = False: If True, boxes are drawn with widths proportional to the square-roots of the number of observations in the groups.
notchwidth : float = 0.5: Width of notch relative to the body width.
fatten : float = 2: A multiplicative factor used to increase the size of the middle bar across the box.

Examples

import numpy as np
import pandas as pd

from plotnine import (
    ggplot,
    aes,
    geom_boxplot,
    geom_jitter,
    scale_x_discrete,
    coord_flip,
)
from plotnine.data import pageviews

A box and whiskers plot

The boxplot compactly displays the distribution of a continuous variable.

	year	month	passengers
0	1949	January	112
1	1949	February	118
2	1949	March	132
3	1949	April	129
4	1949	May	121

A Basic Boxplot

# Gallery, distributions
(
    ggplot(flights)
    + geom_boxplot(aes(x="factor(month)", y="passengers"))
    + scale_x_discrete(labels=months, name="month")  # change ticks labels on OX
)

Horizontal boxplot

(
    ggplot(flights)
    + geom_boxplot(aes(x="factor(month)", y="passengers"))
    + coord_flip()
    + scale_x_discrete(
        labels=months[::-1],
        limits=flights.month[11::-1],
        name="month",
    )
)

Boxplot with jittered points:

(
    ggplot(flights, aes(x="factor(month)", y="passengers"))
    + geom_boxplot()
    + geom_jitter()
    + scale_x_discrete(labels=months, name="month")  # change ticks labels on OX
)

Precomputed boxplots

For datasets that do not fit in memory, you can precompute the boxplot metrics (for example by aggregating the statistics using database queries) and then use geom_boxplot with stat="identity".

# Precompute the metrics
def q25(x):
    return x.quantile(0.25)

def q75(x):
    return x.quantile(0.75)
    
pageviews["hour"] = pageviews.date_hour.dt.hour
precomputed_metrics = pageviews.groupby("hour").agg({'pageviews': ["min", q25, "median", q75, "max"]})
precomputed_metrics.columns = [col_name[1] for col_name in precomputed_metrics.columns]  
precomputed_metrics = precomputed_metrics.reset_index()
precomputed_metrics.head()

	hour	min	q25	median	q75	max
0	0	8437.500380	8842.109077	9297.046035	9600.362430	11762.446233
1	1	8852.123978	9177.938537	9457.821814	10530.072887	11974.437292
2	2	8793.076686	9176.462389	9704.885172	10446.315276	12105.406628
3	3	8683.606449	9574.722286	10615.670464	11290.246605	11651.443193
4	4	8252.974951	9898.998785	10959.909095	11409.657288	11603.711837

(
    ggplot(precomputed_metrics)
    + geom_boxplot(
        aes(x="factor(hour)", ymin="min", lower="q25", middle="median", upper="q75", ymax="max"),
        stat="identity"
    )
)

Source: A box and whiskers plot

Parameters

See Also

Examples

A box and whiskers plot

A Basic Boxplot

Precomputed boxplots