Multi-Plot Grids (FacetGrid, lmplot)

Seaborn lets you create multiple subplots based on categories, which makes it easier to compare patterns across groups.

Two tools are especially helpful for this: FacetGrid and lmplot.

A FacetGrid lays out a grid of subplots using the unique values of one or more categorical variables.

You can define:

• rows: variable that splits plots vertically
• cols: variable that splits plots horizontally
• hue (optional): variable that colors data points

This helps you break a complex dataset into smaller and clearer views.

Copy
import seaborn as sns
import matplotlib.pyplot as plt

tips = sns.load_dataset("tips")

g = sns.FacetGrid(tips, row="sex", col="time", hue="smoker", margin_titles=True)
g.map_dataframe(sns.scatterplot, x="total_bill", y="tip")
g.add_legend()
plt.show()

Copy
g = sns.FacetGrid(tips, col="day")
g.map_dataframe(sns.histplot, x="total_bill", bins=20)
plt.show()

lmplot combines a regression plot with a FacetGrid.

It shows the relationship between variables along with a fitted trend line, and it can split the data by category.

Copy
sns.lmplot(
    data=tips,
    x="total_bill",
    y="tip",
    hue="sex",
    col="time",
    height=4,
    scatter_kws={"alpha": 0.6}
)
plt.show()

• Use FacetGrid when you want a flexible grid and plan to map different plot functions (scatter, hist, KDE, bar).
• Use lmplot when you want regression lines with optional faceting and coloring in one step.

• Limit the number of facets: Too many categories can make the grid hard to read.
• Keep axis ranges consistent: Shared scales make comparisons easier (sharex, sharey).
• Add legends and titles: Use add_legend() and margin_titles=True for clarity.

Now explore these patterns step by step in the notebook on the right side of the screen.