Results table to figure
Skill StaryMoon/results-table-to-figure-skill/skills/results-table-to-figure
Convert CSV or simple LaTeX result tables into editable SVG/PDF and high-resolution PNG bar, line, or heatmap figures with normalized data and provenance. Use when Codex needs to turn benchmark tables, ablations, model comparisons, or experiment summaries into publication-ready plots without retyping values.From its SKILL.md
npx -y skills add StaryMoon/results-table-to-figure-skill --skill results-table-to-figureAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.6 KB, 289 tokens by cl100k_base, as published. Nobody here has run it
Results Table to Figure
Keep the table as the source of truth. Parse values once, write a normalized CSV, then generate all visual formats from that normalized data.
Workflow
- Confirm which column contains methods or x values and which numeric columns should be plotted.
- Run:
python3 scripts/table_to_figure.py results.csv \
--output output/result-figure \
--kind bar \
--x-column method \
--metrics psnr,ssim
- Compare
normalized.csvwith the source before accepting the plot. - Inspect
figure.svgfor labels, clipping, ordering, and misleading scales. - Use SVG or PDF for editable publication workflows; use PNG for previews.
Supported inputs
- CSV with a header row.
- Simple LaTeX
tabularcontent using&columns and\\rows. - Numeric cells such as
30.12,30.12 ± 0.08, and\textbf{30.12}.
Quality rules
- Do not use a truncated y-axis without making the lower bound visible.
- Do not combine metrics with incompatible scales in one bar chart.
- Preserve method order unless the user requests sorting.
- Validate parsed values before discussing winners.
- Keep provenance JSON with publication assets.
See references/plot-selection.md for chart choices.
What ships with it: 3 files
9.8 KB alongside SKILL.md, 1 of them executable
agents/
- openai.yaml241 B
references/
- plot-selection.md497 B
scripts/
- table_to_figure.pyruns9.1 KB