FLoRA Explorer

A static, GitHub-Pages-friendly dashboard for the FORRT Library of Replication Attempts (FLoRA).

Five tabs:

Tab	What it shows	Refreshed
Overview	Headline counts, outcome distribution, About, FAQ, citations	Daily
Browse Studies	Full searchable DataTable + mobile card list	Daily
Years & Disciplines	Year/journal/discipline breakdowns of outcomes	Daily
Citation Impact	OpenCitations event-study of citation changes after replication	Weekly
Mean Citedness	Journal-level OMC vs replication success (R analysis)	Weekly

Every tab shows a "Last updated" stamp pulled from the relevant *_meta.json next to the data.

Contributions of new dashboards/tabs are welcome — see CONTRIBUTING.md.

Layout

.
├── index.html
├── assets/
│   ├── styles.css
│   ├── app.js                 # Overview / Browse / Years & Disciplines / Mean Citedness loader
│   ├── citation-impact.js     # Citation Impact tab (lazy-loaded)
│   └── logo.svg
├── data/                      # All written by GitHub Actions, except disciplines.json
│   ├── disciplines.json       # Hand-curated journal → discipline map (commit-controlled)
│   ├── flora.csv              # Daily snapshot of upstream flora.csv
│   ├── flora_meta.json
│   ├── flora_with_omc.csv     # FLoRA + OpenAlex OMC per journal_o (weekly)
│   ├── flora_with_omc_meta.json
│   ├── impact_factor_data.json # Chart-ready Mean Citedness data (weekly)
│   ├── impact_factor_meta.json
│   ├── author_overlap_data.json # Authorship Overlap tab data (daily)
│   ├── author_overlap_meta.json
│   ├── meta.json              # Citation pipeline (weekly)
│   ├── aggregate.json
│   └── originals.json
├── scripts/
│   ├── refresh_flora.py       # Daily flora.csv snapshot
│   ├── refresh_data.py        # Weekly OpenCitations citation pipeline
│   ├── compute_omc.py         # Weekly OpenAlex OMC enrichment
│   ├── compute_author_overlap.py # Daily authorship-overlap computation
│   ├── render_impact_factor.R # Computes Mean Citedness stats, writes JSON directly
│   ├── run_fect.R             # ETWFE overlay for Citation Impact (not yet wired into a workflow)
│   └── requirements.txt
├── archive/                   # Superseded scripts/outputs, kept for reference only
│   ├── scripts/                #   compute_impact_json.py, impact_factor.Rmd
│   └── data/                   #   impact_factor.html + impact_factor_figs/
├── cache/                     # API caches committed between runs
│   ├── oc/                    # OpenCitations
│   └── openalex_venues.json   # OpenAlex sources
└── .github/workflows/
    ├── refresh-flora.yml          # Daily   03:00 UTC (flora.csv + author overlap)
    ├── refresh-data.yml           # Weekly Mon 04:00 UTC (citation pipeline)
    ├── refresh-impact-factor.yml  # Weekly Mon 05:00 UTC (OMC + R render)
    └── clean-json.yml             # Manual maintenance helper

Deploying on GitHub Pages

1. Push this repository to GitHub.
2. Settings → Pages → Build and deployment → Source: Deploy from a branch. Pick main and / (root).
3. Settings → Secrets and variables → Actions — add the secrets used by the data-refresh workflows:
   • MY_EMAIL — your email. Used in the polite User-Agent header for OpenAlex and OpenCitations. Required.
   • OC_API_KEY — optional OpenCitations API key (raises rate limits).
4. Settings → Actions → General → Workflow permissions: select Read and write permissions so the bot can commit refreshed data back to the repo.
5. Trigger the workflows manually the first time (Actions → Refresh FLoRA snapshot → Run workflow, etc.) so the data files appear. Subsequent runs follow the cron schedule.

That is the entire deploy — no build step, no server.

Until the first refresh has run, the Explorer falls back to fetching flora.csv directly from the FReD-data repository so the Overview / Browse / Years tabs still work.

Running data refreshes locally

pip install -r scripts/requirements.txt

# Daily snapshot (just pulls flora.csv)
python scripts/refresh_flora.py

# Mean Citedness pipeline (needs R)
MY_EMAIL=you@example.org python scripts/compute_omc.py
Rscript scripts/render_impact_factor.R

# Citation Impact pipeline (long-running; uses OpenCitations)
MY_EMAIL=you@example.org python scripts/refresh_data.py

# Authorship Overlap (needs flora.csv already downloaded)
python scripts/compute_author_overlap.py

R packages required: jsonlite, mgcv.

Editing the disciplines map

data/disciplines.json is the single source of truth for the journal → discipline mapping used both by the JS frontend (Years & Disciplines tab) and by the R Mean Citedness analysis. Edit it once and both views update on the next deploy / refresh.

Acknowledgements

• Data: FORRT FReD project
• Citation backend: OpenCitations COCI
• Journal Mean Citedness: OpenAlex Sources API

Suggested citation

Wallrich, L., & Röseler, L. (2026). FLoRA Explorer [Website]. https://forrt.org/flora-explorer/