# Matplotlib

> Matplotlib OO/pyplot guidance for custom plots via agent_generated_code. Use for fine-grained control. Prefer nature_figure for manuscript figures and seaborn for quick statistical EDA.

- Skill: `ai4protein/matplotlib` (Agent Skill, multi-file: 6 files)
- Install (CLI): `npx skillmds@latest add ai4protein/matplotlib`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ai4protein/matplotlib/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- License: https://github.com/matplotlib/matplotlib/tree/main/LICENSE
- Author: ai4protein (https://skillmd.com/u/ai4protein)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/ai4protein/matplotlib

---


# Matplotlib

## Overview

Foundational plotting. In VenusFactory execute via **`agent_generated_code`**. Helpers: `src/tools/visualize/matplotlib/plot_template.py`, `style_configurator.py`.

## When to use / NOT

| Use matplotlib | Prefer |
|----------------|--------|
| Custom artists, insets, unusual projections | — |
| Statistical EDA defaults | `seaborn` |
| Nature submission figures | `nature_figure` (load first) |

## Quick patterns

```python
import matplotlib.pyplot as plt
fig, ax = plt.subplots(figsize=(5, 4))
ax.plot(x, y)
ax.set_xlabel("..."); ax.set_ylabel("...")
fig.savefig(out_png, dpi=300, bbox_inches="tight")
plt.close(fig)
```

Prefer the **OO API** (`fig, ax = plt.subplots`) over pyplot state for multi-step agent code.

## Common mistakes

- Forgetting dpi≥300 / session `output_dir`
- Leaving interactive `plt.show()` in headless runs
- Skipping `nature_figure` for publication output

## References (progressive disclosure)

**Trust order:** `SKILL.md` (hub tools & envelopes) → topic refs → `references/legacy_guide.md` (archived; may be outdated).

```text
read_skill(skill_id="matplotlib", relative_path="references/legacy_guide.md")
```

Load legacy only after the hub workflow in this file is insufficient.