Skip to content

Generate reports

After running the pipeline and analysis, generate an interactive HTML report covering all studies with available data:

Terminal window
just report

Output: data/report/index.html and individual study pages.

Preview the report in your browser:

Terminal window
just serve-report

Opens at http://localhost:8000. To use a custom port:

Terminal window
just serve-report 9000

The report generator needs:

  1. Quality measurements (data/metrics/<study-id>/quality.json) — from running the pipeline
  2. Analysis outputs (data/analysis/<study-id>/) — from analyzing results
  3. Comparison images (optional) — from generating comparisons

Each study page includes:

  • Interactive Plotly visualizations — rate-distortion curves, quality-vs-parameter plots, efficiency plots with zoom, pan, and hover tooltips
  • Static SVG figures — downloadable vector plots from the analysis step
  • CSV statistics — downloadable per-study statistics tables
  • Comparison figures — visual comparison grids with matched quality and matched file-size figure groups (if generated); each raster figure opens in a PhotoSwipe lightbox for full-resolution zoom and swipe navigation
  • Tool versions — encoder and measurement tool versions for reproducibility

For crop-impact studies, the report also includes annotated originals that show the selected comparison fragment, the measured analysis fragment, and the crop windows used for the different crop levels.

Each report page has:

  • A table of contents at the top with expandable sub-entries for each comparison figure group (matched quality vs. matched file size).
  • Back to top links throughout the page so the reader can quickly return to the table of contents from any section.
data/report/
├── index.html # Landing page with study list
├── format-comparison.html # Individual study page
├── ...
├── data/ # Downloadable data per study
│ └── <study-id>/
│ ├── *_statistics.csv
│ └── *.svg
└── assets/
└── plotly-basic.min.js # Plotly library (auto-downloaded)
  • “No quality.json found”: Run the pipeline first — just pipeline <study-id> <time>
  • “No static figures found”: Run analysis — just analyze <study-id>
  • Plotly not interactive: Regenerate the report — just report re-downloads the Plotly bundle