This project provides a Python backend for the Chicago Crimes dataset task.

It covers:

• dataset structure understanding via metadata API
• relational storage workflow (raw table to clean typed table)
• at least 10 independent SQL analysis angles
• frontend-ready API contract for charts and text conclusions

crime_analytics_backend/
  app/
    core/
      config.py
    routers/
      analysis.py
      metadata.py
      system.py
    services/
      analysis_service.py
    static/
      app.js
      dashboard_snapshot.json
      index.html
    db.py
    main.py
    schemas.py
  docs/
    API.md
  scripts/
    build_dashboard_snapshot.py
    opengauss_import.py
    opengauss_prepare_clean.py
  sql/
    prepare_clean_table.sql
    analysis_queries.sql
  requirements.txt
  start_server.py
  .env.example

• Python 3.10+
• Huawei Cloud openGauss
• Existing raw import table: crimes_raw

After cloning, you can run directly:

cd d:/database/crime_analytics_backend
python start_server.py --host 127.0.0.1 --port 8016

The project will auto-install dependencies and use built-in openGauss defaults.

cd d:/database/crime_analytics_backend
C:/Users/73110/AppData/Local/Microsoft/WindowsApps/python3.11.exe -m pip install -r requirements.txt

The project now provides an auto-bootstrap launcher. If dependencies are missing or version-mismatched, it will install them automatically on startup. So manual install can be skipped.

Optional: copy .env.example to .env if you want to override default openGauss connection settings.

cd d:/database/crime_analytics_backend
python scripts/opengauss_prepare_clean.py

cd d:/database/crime_analytics_backend
python start_server.py --host 127.0.0.1 --port 8016

After startup, browser auto-opens /dashboard by default.

If you want to disable auto-open:

python start_server.py --host 127.0.0.1 --port 8016 --no-open-browser

Service URLs:

• API docs: http://127.0.0.1:8016/docs
• Dashboard: http://127.0.0.1:8016/dashboard

Stop service with Ctrl + C in the same terminal.

When data volume is very large, default dashboard queries can be slow. You can pre-build a static snapshot so the first screen loads from local JSON instead of querying all analysis endpoints.

cd d:/database/crime_analytics_backend
python scripts/build_dashboard_snapshot.py

This generates:

app/static/dashboard_snapshot.json

Dashboard behavior after this change:

• default view (no filters + topN=10): uses static snapshot for fast loading
• filtered view (year/type changed): still uses live API queries

If database is unavailable, you can generate a sample snapshot:

python scripts/build_dashboard_snapshot.py --sample

1. Fill OG_PASSWORD in .env.
2. Run importer script:

cd d:/database/crime_analytics_backend
python scripts/opengauss_import.py

Then build analysis table (crimes_clean) used by API:

python scripts/opengauss_prepare_clean.py

If the user does not have permission on public schema, use:

python scripts/opengauss_import.py --schema testuser

Quick verification with partial data:

python scripts/opengauss_import.py --schema testuser --max-rows 10000

The script will:

• connect to openGauss server
• create schema/table if not exists
• truncate target table by default
• bulk import CSV using COPY

Optional arguments:

• --no-truncate: keep existing rows and append
• --table crimes_raw_new: import to another table name
• --max-rows 10000: import only first N rows for connectivity check
• --password your_password: pass password from CLI (less secure than .env)

• Swagger UI: http://127.0.0.1:8016/docs
• ReDoc: http://127.0.0.1:8016/redoc
• Frontend integration contract: docs/API.md

1. Dashboard bundle (recommended for frontend dynamic refresh)
2. Annual trend
3. Weekly distribution
4. Hourly distribution
5. Crime type share
6. District comparison
7. Day-hour heatmap
8. YoY trend of top crime types
9. Text conclusions bundle (10+ statements)

• sql/analysis_queries.sql contains standalone SQL examples for report writing.
• Most analysis and metadata APIs support sample=true for frontend demo without database dependency.
• start_server.py checks dependency versions against requirements.txt and auto-installs missing items before server startup.
• For production deployment, add authentication, tighter CORS, and caching.