installation.md

Advanced installation

There are three different installation methods with various levels of containerization:

• 📦 Fully containerized API and database (recommended)
• 💻 Containerized database, local python (recommended for development)
• 🤓 Fully local, not using docker at all (if you already have postgres)

Fully containerized API and database

1. First, clone this repository and start a terminal
2. Then, run the docker containers in this network, exposing the API on port 8000.

   docker compose up

3. Navigate to localhost:8000/recommend/1?n=5 or see the docs at localhost:8000/docs

Dockerized database

1. Install Python
2. Clone this repo and cd to it
3. Build and run the database dockerfile

   docker build . -f dockerfiles/pagiladb.dockerfile -t pagiladb
   docker run -p 5432:5432 -e POSTGRES_DB=pagila -e POSTGRES_PASSWORD=postgres pagiladb

4. Install python requirements:

   pip install -r requirements.txt

5. Run the recommender API via the uvicorn python package:

   uvicorn film_api:app

6. Navigate to localhost:8000/recommend/1?n=5. See the docs at localhost:8000/docs

Fully local installation

1. Install PostgreSQL & python
2. Clone this repo and cd to it.
3. Install and generate the example database

   git clone https://github.com/xzilla/pagila.git
   mv pagila/pagila-schema.sql pagila/1_schema.sql
   mv pagila/pagila-data.sql pagila/2_data.sql
   rm pagila/pagila-insert-data.sql pagila/README
   createdb -U postgres pagila
   psql -U postgres -d pagila -f pagila/1_schema.sql
   psql -U postgres -d pagila -f pagila/2_data.sql
   psql -U postgres -d pagila -c "CREATE extension tablefunc;"

4. Install python requirements:

   pip install -r requirements.txt

5. Run the recommender API via the uvicorn python package:

   uvicorn film_recommender:app

6. Navigate to localhost:8000/recommend/1?n=5. See the docs at localhost:8000/docs