coralfuture.org App

The website is available at coralfuture.org, run by Voolstra lab. The motivation is to build a global database of standardized thermal tolerance ED50 values as determined by CBASS to enable meta-analyses and -comparisons.

Deploy

1. docker compose up -d
2. Frontend:

   cd react_app
   npm install
   npm run build
   rsync -a build/ /var/www/hemorrhagia.online/
   

3. Django superuser (when needed):

   docker compose exec django-app python manage.py createsuperuser
   

4. Optional seed users: edit django_app/user_data.example.json, then save it as user_data.json in the same folder before docker compose up -d—startup scripts will create those accounts automatically.
5. After code updates:

   docker compose exec django-app python manage.py migrate
   docker compose exec django-app python manage.py collectstatic --noinput
   sudo nginx -t && sudo systemctl reload nginx
   

API endpoints

• GET / – FastAPI ED calculator home page with upload form.
• POST /process – handles file upload/example data, invokes the R workflow, returns table + plots.
• GET /download-csv – downloads the latest ED table (used by the web UI button).

FastAPI ED calculator

1. UI posts CSV/XLSX or toggled example data to /process.
2. FastAPI writes the dataset to a temp CSV and calls calculate_eds.R.
3. The R script loads CBASSED50, preprocesses, fits DRMs with is_curveid = TRUE, and writes ED5/50/95 plus PNG plots.
4. FastAPI reads the output CSV and PNGs, embeds the table + base64 images into the results template with download buttons.

Django backend endpoints

• / — legacy Django home, keeps basic landing view, redirects to React when needed.
• /admin/ — standard Django admin.
• /projects/ — SSR views for project listings (see projects.urls).
• /user/ — auth/profile views (see users.urls).
• /api/auth/... — authenticated REST endpoints: cart (/cart/, /cart/group/<id>/, /cart/export/), session helpers (/status/, /csrf/, /login/, /logout/), CSV upload/check/ED50 calculation (/upload-csv/, /check-csv-ed50/, /calculate-ed50/).
• /api/public/... — read-only data feeds: statistics/, biosamples/, colonies/, observations/, projects/, projects/<id>/, and thermal layers (thermal-tolerances/, thermal-tolerances/max-min/, breakpoint-temperatures/, breakpoint-temperatures/max-min/, thermal-limits/, thermal-limits/max-min/).

Ports and services

Service	Port	Notes
Django backend	8000	API + admin, proxied by nginx /api, /admin.
FastAPI ED50 app	8001	/shiny + /process, runs the R workflow.
React build (nginx)	443	Deployed to /var/www/hemorrhagia.online.

Deployment cheatsheet

1. Build frontend

   cd react_app
   npm install
   npm run build
   sudo cp -r build/* /var/www/coralfuture.org/

2. Rebuild backend containers

   docker compose build
   docker compose up -d

3. Static files / migrations (when needed)

   docker compose exec django-app python manage.py migrate
   docker compose exec django-app python manage.py collectstatic --noinput

4. Reload nginx

   sudo nginx -t && sudo systemctl reload nginx

Django data model quick-notes

• users app: wraps Django auth, adds profile info, and exposes /api/auth/... endpoints for login/logout, cart management, CSV uploads, and ED50 calculations. Superusers manage accounts via /admin/.
• projects app: handles studies (projects) and associated colonies/observations. React pulls these via /api/public/projects and details via /api/public/projects/<id>/.
• “Studies”/“cart” flow: authenticated users build a cart of colonies (UserCart, CartGroup) through the API, then export ED tables via /api/auth/cart/export/. Upload endpoints (upload-csv, check-csv-ed50, calculate-ed50) persist uploads for review.