python-dataviz

# Python Data Visualization

Python data visualization with matplotlib and seaborn for creating publication-quality figures, statistical graphics, and exploratory visualizations.

## When to use each library

**Matplotlib** is the foundational plotting library. Use it for:

- Fine-grained control over every plot element
- Custom layouts with GridSpec or subplot_mosaic
- 3D visualizations
- Animations
- Embedding plots in GUI applications
- When you need low-level customization

**Seaborn** builds on matplotlib for statistical visualization. Use it for:

- Statistical plots with automatic aggregation and confidence intervals
- Dataset-oriented plotting from DataFrames
- Faceted multi-panel figures (small multiples)
- Distribution visualization (KDE, histograms, violin plots)
- Correlation matrices and clustered heatmaps
- Publication-ready aesthetics with minimal code

**Combined approach**: Use seaborn for the main visualization, then customize with matplotlib.

## Core concepts

### Matplotlib hierarchy

1. **Figure** - Top-level container for all plot elements
2. **Axes** - Actual plotting area (one Figure can have multiple Axes)
3. **Artist** - Everything visible (lines, text, ticks, patches)
4. **Axis** - The x/y number lines with ticks and labels

### Two matplotlib interfaces

**Object-oriented interface (recommended)**:

```python
import matplotlib.pyplot as plt

fig, ax = plt.subplots(figsize=(10, 6))
ax.plot(x, y, linewidth=2, label='data')
ax.set_xlabel('X Label')
ax.set_ylabel('Y Label')
ax.legend()
plt.savefig('figure.png', dpi=300, bbox_inches='tight')
```

**pyplot interface** (quick exploration only):

```python
plt.plot(x, y)
plt.xlabel('X Label')
plt.show()
```

Always use the object-oriented interface for production code.

### Seaborn function levels

**Axes-level functions** plot to a single matplotlib Axes:

- Accept `ax=` parameter for placement
- Return Axes object
- Examples: `scatterplot`, `histplot`, `boxplot`, `heatmap`

**Figure-level func