seaborn

Compute the correlation matrix with df.corr(), then pass it to sns.heatmap with annot=True, fmt='.2f', cmap='coolwarm', and center=0. Use numpy's triu to mask the upper triangle for a cleaner display.

Axes-level functions like scatterplot and boxplot draw onto a single matplotlib Axes and accept an ax parameter. Figure-level functions like relplot, displot, and catplot manage entire figures with built-in faceting via col and row parameters.

Use boxplot for quartile summaries, violinplot to show full distribution shape with KDE, and swarmplot to display every individual point without overlap. Swarmplot becomes slow beyond roughly 1000 points, so use stripplot for larger datasets.

Yes, seaborn is designed around pandas DataFrames in long-form (tidy) format where each variable is a column. Passing data with named x, y, and hue columns automatically produces labeled axes and legends.

The KDE bandwidth controls smoothness and is adjusted with the bw_adjust parameter. Values below 1 produce more jagged curves that follow the data closely, while values above 1 produce smoother curves.

Use seaborn.objects for complex layered visualizations, programmatic plot generation, and fine-grained control over marks, stats, and scales. The function interface remains simpler for quick single-purpose exploratory plots.