haipipe-display-figure

name	haipipe-display-figure
description	Generate publication-quality data plots from experiment results (line/bar/scatter/heatmap/box). Use when user says "画图", "作图", "generate figures", "paper plots", or needs data-driven plots for a paper. The plot renderer of the display family; tables are rendered by haipipe-display-table.
argument-hint	[figure-plan-or-data-path]
allowed-tools	Bash(*), Read, Write, Edit, Grep, Glob, Agent, mcp__codex__codex, mcp__codex__codex-reply
metadata	{"version":"0.2.1","last_updated":"2026-07-27","summary":"Generate publication-quality data plots from a provenance-bound Display Intake."}

Paper Display Figure: Publication-Quality Plots from Experiment Data

Generate the data plots for a paper based on: $ARGUMENTS

Boundary: this skill renders plots only. For any other display kind (tables, diagrams, AI concept art), see the sibling-routing table in ../../ref/display-unit-output-contract.md.

Output: write into a display unit

The plot goes into a displays/displayNN-<slug>/ unit per the shared contract: ../../ref/display-unit-output-contract.md. THIS renderer's row: asset -> assets/figure.pdf; rebuild spec -> recipe/gen_*.py (+ recipe/paper_plot_style.py).

For a new unit, read intake/manifest.yaml before doing anything else. The plot script reads only the manifest's approved intake/inputs/ snapshot. It never searches a task folder, re-derives values, or chooses rows from an arbitrary result file. Legacy source/ units remain valid only through the compatibility path in the shared contract.

Scope: What This Skill Can and Cannot Do

Category	Can auto-generate?	Examples
Data-driven plots	✅ Yes	Line plots (training curves), bar charts (method comparison), scatter plots, heatmaps, box/violin plots
Comparison tables	➡️ Use haipipe-display-table	LaTeX tables (prior bounds, method features, ablation) now live in the dedicated table renderer
Multi-panel figures	✅ Yes	Subfigure grids combining multiple plots (e.g., 3×3 dataset × method)
Architecture/pipeline diagrams	❌ No — manual	Model architecture, data flow diagrams, system overviews. At best can generate a rough TikZ skeleton, but expect to draw these yourself using tools like draw.io, Figma, or TikZ
Generated image grids	❌ No — manual	Grids of generated samples (e.g., GAN/diffusion outputs). These come from running your model, not from this skill
Photographs / screenshots	❌ No — manual	Real-world images, UI screenshots, qualitative examples

In practice: For a typical ML paper, this skill handles the data plots (a large share of the figure set). Tables go to haipipe-display-table; the hero figure / architecture diagram / qualitative results are created via the diagram/illustration skills or manually and placed in figures/ before running /haipipe-paper section-edit. The skill will detect manually-made figures as "existing figures" and preserve them.

Constants

• STYLE = publication — Visual style preset. Options: publication (default, clean for print), poster (larger fonts), slide (bold colors)
• DPI = 300 — Output resolution
• FORMAT = pdf — Output format. Options: pdf (vector, best for LaTeX), png (raster fallback)
• COLOR_PALETTE = tab10 — Default matplotlib color cycle. Options: tab10, Set2, colorblind (deuteranopia-safe)
• FONT_SIZE = 10 — Base font size (matches typical conference body text)
• FIG_DIR — for a paper, the display unit displays/displayNN-slug/ (plot -> assets/figure.pdf, scripts -> recipe/). Flat figures/ only with no paper.
• REVIEWER_MODEL = gpt-5.5 — Model used via Codex MCP for figure quality review.

Inputs

1. Display contract — the unit's README.md and the paper-stage brief.
2. Display Intake — intake/manifest.yaml and its approved CSV/JSON snapshot.
3. Existing candidate — only when the caller asked to refine that named candidate.

If no display unit or verified Intake exists, stop and ask the caller to create one.

Workflow

Step 1: Read the Unit Brief and Intake

Read the unit README.md for the claim, audience, caption intent, and target section. Then read intake/manifest.yaml and verify the declared snapshot hash before plotting. The Display stage, not this renderer, already decided the figure plan and form.

If the manifest has no role: values source, stop and route a concept visual to the diagram or illustration renderer instead.

Step 2: Set Up Plotting Environment

Create a shared style configuration script:

# paper_plot_style.py — shared across all figure scripts
import matplotlib.pyplot as plt
import matplotlib
matplotlib.rcParams.update({
    'font.size': FONT_SIZE,
    'font.family': 'serif',
    'font.serif': ['Times New Roman', 'Times', 'DejaVu Serif'],
    'axes.labelsize': FONT_SIZE,
    'axes.titlesize': FONT_SIZE + 1,
    'xtick.labelsize': FONT_SIZE - 1,
    'ytick.labelsize': FONT_SIZE - 1,
    'legend.fontsize': FONT_SIZE - 1,
    'figure.dpi': DPI,
    'savefig.dpi': DPI,
    'savefig.bbox': 'tight',
    'savefig.pad_inches': 0.05,
    'axes.grid': False,
    'axes.spines.top': False,
    'axes.spines.right': False,
    'text.usetex': False,  # set True if LaTeX is available
    'mathtext.fontset': 'stix',
})

# Color palette
COLORS = plt.cm.tab10.colors  # or Set2, or colorblind-safe

def save_fig(fig, name, fmt=FORMAT):
    """Save figure to FIG_DIR with consistent naming."""
    fig.savefig(f'{FIG_DIR}/{name}.{fmt}')
    ()

Step 3: Auto-Select Figure Type

Use this decision tree for data-driven figures (inspired by Imbad0202/academic-research-skills):

Data Pattern	Recommended Type	Size
X=time/steps, Y=metric	Line plot	0.48\textwidth
Methods × 1 metric	Bar chart	0.48\textwidth
Methods × multiple metrics	Grouped bar / radar	0.95\textwidth
Two continuous variables	Scatter plot	0.48\textwidth
Matrix / grid values	Heatmap	0.48\textwidth
Distribution comparison	Box/violin plot	0.48\textwidth
Multi-dataset results	Multi-panel (subfigure)	0.95\textwidth
Prior work comparison / coefficients	(table) → use haipipe-display-table	—

Step 4: Generate Each Figure

For the current display unit, create a standalone Python script in recipe/:

Line plots (training curves, scaling):

# gen_fig2_training_curves.py
from paper_plot_style import *
import json

with open('intake/inputs/exp_results.json') as f:
    data = json.load(f)

fig, ax = plt.subplots(1, 1, figsize=(5, 3.5))
ax.plot(data['steps'], data['fac_loss'], label='Factorized', color=COLORS[0])
ax.plot(data['steps'], data['crf_loss'], label='CRF-LR', color=COLORS[1])
ax.set_xlabel('Training Steps')
ax.set_ylabel('Cross-Entropy Loss')
ax.legend(frameon=False)
save_fig(fig, 'fig2_training_curves')

Bar charts (comparison, ablation):

from paper_plot_style import *
import pandas as pd

data = pd.read_csv('intake/inputs/comparison.csv')
fig, ax = plt.subplots(1, 1, figsize=(5, 3))
bars = ax.bar(data['method'], data['value'], color=[COLORS[i] for i in range(len(data))])
ax.set_ylabel('Accuracy (%)')
# Add value labels on bars
for bar, val in zip(bars, data['value']):
    ax.text(bar.get_x() + bar.get_width()/2, bar.get_height() + 0.3,
            f'{val:.1f}', ha='center', va='bottom', fontsize=FONT_SIZE-1)
save_fig(fig, 'fig3_comparison')

Comparison / coefficient tables (LaTeX): out of scope — use haipipe-display-table, which owns booktabs rules, significance stars, SE rows, panels, and table notes. Do not emit .tex tables from this skill.

Architecture/pipeline diagrams are outside this skill's scope. Route them to haipipe-display-diagram or haipipe-display-illustration through the Display stage.

Step 5: Run All Scripts

# Run all figure generation scripts
for script in gen_fig*.py; do
    python "$script"
done

Verify all output files exist and are non-empty.

Step 6: Hand Back to the Unit Wrapper

The renderer writes the asset and recipe only. float.tex is caller-owned: after the Paper adapter supplies an approved caption, label, and placement, a renderer may refresh just its asset reference under the shared contract. It never invents or changes those semantic fields. Do not create a parallel latex_includes.tex file or write an ad hoc figure block in a section.

The Paper adapter places the accepted unit through its existing float.tex.

Step 7: Figure Quality Review with REVIEWER_MODEL

Send figure descriptions and captions to GPT-5.5 for review:

mcp__codex__codex:
  model: gpt-5.5
  config: {"model_reasoning_effort": "xhigh"}
  prompt: |
    Review these figure/table plans for a [VENUE] submission.

    For each figure:
    1. Is the caption informative and self-contained?
    2. Does the figure type match the data being shown?
    3. Is the comparison fair and clear?
    4. Any missing baselines or ablations?
    5. Would a different visualization be more effective?

    [list all figures with captions and descriptions]

Step 8: Quality Checklist

Before finishing, verify each figure (from pedrohcgs/claude-code-my-workflow):

• Font size readable at printed paper size (not too small)
• Colors distinguishable in grayscale (print-friendly)
• No title inside figures — titles go only in LaTeX \caption{} (from pedrohcgs)
• Legend does not overlap data
• Axis labels have units where applicable
• Axis labels are publication-quality (not variable names like emp_rate)
• Figure width fits single column (0.48\textwidth) or full width (0.95\textwidth)
• PDF output is vector (not rasterized text)
• No matplotlib default title (remove plt.title for publications)
• Serif font matches paper body text (Times / Computer Modern)
• Colorblind-accessible (if using colorblind palette)

Output

The display unit layout (approved values -> intake/inputs/, asset -> assets/figure.pdf, rebuild recipe -> recipe/gen_figNN_*.py + recipe/paper_plot_style.py) and the no-paper flat fallback are the shared contract: ../../ref/display-unit-output-contract.md.

Figure Type Reference

Type	When to Use	Typical Size
Line plot	Training curves, scaling trends	0.48\textwidth
Bar chart	Method comparison, ablation	0.48\textwidth
Grouped bar	Multi-metric comparison	0.95\textwidth
Scatter plot	Correlation analysis	0.48\textwidth
Heatmap	Attention, confusion matrix	0.48\textwidth
Box/violin	Distribution comparison	0.48\textwidth
Architecture	System overview	0.95\textwidth
Multi-panel	Combined results (subfigures)	0.95\textwidth