RentTheRunway Recommendations

This tutorial demonstrates how to configure a recommendation engine using the RentTheRunway dataset. The dataset contains 192,544 ratings from 105,508 users on 5,850 items. This example uses the local table connector; the same approach applies to other supported connectors.

Accompanying notebook

CLI Setup

Install the CLI

pip install shaped

info

Shaped supports Python 3.8 to 3.11. See installation instructions if you need to install pip.

Initialize the CLI

shaped init --api-key <YOUR_API_KEY>

If you don't have an API key, see How to get an API key.

Data Preparation

Download the dataset

CLI

wget https://mcauleylab.ucsd.edu/public_datasets/data/renttherunway/renttherunway_final_data.json.gz --no-check-certificate
gunzip renttherunway_final_data.json.gz

The dataset contains a single JSON file (renttherunway_final_data.json) with events, users, and items combined. Convert to CSV and preprocess:

• Convert weight and height columns to numerical types
• Convert ratings (1-10 scale) to binary: values > 8 become 1, values ≤ 8 become 0

See the notebook for preprocessing code.

Save as TSV:

data.to_csv('notebook_assets/events.csv', sep='\t', index=False)

This example uses event data along with user and item information. While these are typically separate tables, this dataset combines them. The events.csv file contains events, users, and items. The engine configuration extracts each type using SQL queries.

Create the table

Define a custom table schema:

events_table_schema.yaml

column_schema:
  age: Int32
  body_type: String
  bust_size: String
  category: String
  fit: String
  height: Int32
  item_id: String
  rating: Int32
  rented_for: String
  review_date: DateTime
  review_summary: String
  review_text: String
  size: Int32
  user_id: String
  weight: Int32
name: rent_runway_events
schema_type: CUSTOM

Create the table and insert data:

CLI

shaped create-table --file notebook_assets/events_table_schema.yaml
shaped table-insert --table-name rent_runway_events --file notebook_assets/events.csv --type 'tsv'

Records upload in batches of 1000. Wait until all 192,544 records are uploaded.

Create the engine

This example uses events, users, and items from a single table. The engine configuration uses SQL queries to extract each type:

rent_runway_recommendations.yaml

data:
  item_table:
    type: query
    query: |
      SELECT item_id, fit, category, size FROM rent_runway_events
  user_table:
    type: query
    query: |
      SELECT user_id, bust_size, weight, body_type, height, age FROM rent_runway_events
  interaction_table:
    type: query
    query: |
      SELECT user_id, item_id, review_date AS created_at, rating AS label, rented_for,
        review_text, review_summary FROM rent_runway_events
training:
  models:
    - name: als
      policy_type: als

Create the engine:

shaped create-engine --file notebook_assets/rent_runway_recommendations.yaml

For details on engine configuration, see Engines documentation.

Monitor engine status

Engine creation and training can take several hours, depending on data volume and attributes. Check status:

shaped list-engines

Response:

[
    "engines": {
        "created_at": "2023-03-18T19:17:51 UTC",
        "engine_name": "rent_runway_recommendations",
        "engine_uri": "https://api.shaped.ai/v2/engines/rent_runway_recommendations",
        "status": "FETCHING",
    }
]

The engine progresses through these stages:

1. SCHEDULING
2. FETCHING
3. TRAINING
4. DEPLOYING
5. ACTIVE

Once the status is ACTIVE, the engine is ready for queries.

Query recommendations

Query recommendations using the Query endpoint. Provide a user_id and the number of results to return.

Using the CLI:

shaped query --engine-name rent_runway_recommendations \
  --query "SELECT * FROM similarity(embedding_ref='als', limit=50, encoder='precomputed_user', input_user_id='\$user_id') LIMIT 5" \
  --parameters '{"user_id": "1"}'

Response:

{
   "results": [
      {
         "id": "427010",
         "score": 0.9
      },
      {
         "id": "182094",
         "score": 0.8
      },
      {
         "id": "332874",
         "score": 0.7
      },
      {
         "id": "827918",
         "score": 0.3
      },
      {
         "id": "403528",
         "score": 0.2
      }
   ]
}

The response contains an array of result objects with item IDs and scores.

Using the REST API:

curl https://api.shaped.ai/v2/engines/rent_runway_recommendations/query \
  -H "x-api-key: <API_KEY>" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "SELECT * FROM similarity(embedding_ref=''als'', limit=50, encoder=''precomputed_user'', input_user_id=''$user_id'') LIMIT 5",
    "parameters": {
      "user_id": "1"
    }
  }'

Clean up

Delete the engine when finished:

shaped delete-engine --engine-name rent_runway_recommendations