academic-plotting

1# Academic Plotting for ML Papers
2 
3Generate publication-quality figures for ML/AI conference papers. Two distinct workflows:
4 
51. **Diagram figures** (architecture, system design, workflows, pipelines) — AI image generation via Gemini
62. **Data figures** (line charts, bar charts, scatter plots, heatmaps, ablations) — matplotlib/seaborn
7 
8## When to Use Which Workflow
9 
10| Figure Type | Tool | Why |
11|-------------|------|-----|
12| Architecture / system diagram | Gemini (Workflow 1) | Complex spatial layouts with boxes, arrows, labels |
13| Workflow / pipeline / lifecycle | Gemini (Workflow 1) | Multi-step processes with connections |
14| Bar chart, line plot, scatter | matplotlib (Workflow 2) | Precise numerical data, reproducible |
15| Heatmap, confusion matrix | matplotlib/seaborn (Workflow 2) | Structured grid data |
16| Ablation table as chart | matplotlib (Workflow 2) | Grouped bars or line comparisons |
17| Pie / donut chart | matplotlib (Workflow 2) | Proportional data (use sparingly in ML papers) |
18| Training curves | matplotlib (Workflow 2) | Loss/accuracy over steps/epochs |
19 
20**Rule of thumb**: If the figure has numerical axes, use matplotlib. If the figure has boxes and arrows, use Gemini.
21 
22---
23 
24## Step 0: Context Analysis & Extraction
25 
26The user will typically provide one of these inputs — not a ready-made specification:
27 
28| Input Type | Example | What to Extract |
29|-----------|---------|-----------------|
30…