Basic Barplot with plotnine

This vignette shows how to create a basic barplot using plotnine. A barplot displays the relationship between a numeric variable and a categorical variable.

This example is a plotnine port of the Basic Barplot tutorial from the Python Graph Gallery, which originally uses matplotlib.

Libraries & Dataset

We use pandas to hold the data and plotnine_extra (which re-exports the full plotnine API) for plotting.

import pandas as pd
from plotnine import (
    ggplot,
    aes,
    geom_col,
    labs,
    theme_minimal,
)

# Create the dataset (same data as the original matplotlib example)
df = pd.DataFrame({
    "category": ["A", "B", "C", "D", "E"],
    "height": [3, 12, 5, 18, 45],
})

Basic Barplot

With plotnine we build plots using the grammar of graphics:

1. ``ggplot(df, aes(…))`` – bind the data and map columns to aesthetics.

2. ``geom_col()`` – draw a bar whose height equals the value in the data (use geom_bar() when you want plotnine to count rows for you).

(
    ggplot(df, aes(x="category", y="height"))
    + geom_col()
)

Customising the plot

plotnine makes it easy to polish the appearance. Below we add a colour fill, labels, and a cleaner theme.

(
    ggplot(df, aes(x="category", y="height"))
    + geom_col(fill="#69b3a2")
    + labs(
        title="Basic Barplot with plotnine",
        x="Category",
        y="Value",
    )
    + theme_minimal()
)

Going further

Because plotnine implements the grammar of graphics, extending this basic example is straightforward:

• Map fill to a column to get grouped / stacked bars.

• Use coord_flip() for horizontal bars.

• Add error bars with geom_errorbar().

• Facet with facet_wrap() or facet_grid().

See the plotnine documentation and the plotnine-extra API reference for more details.