query-model-joined-with-data.mdx

title	Get Batch Predictions
openapi	POST /api/projects/{projectName}/models/{modelName}/predict
sidebarTitle	Get Batch Predictions

There are two ways to get batch predictions:

1. Join the model with data tables and use the query endpoint to query for batch predictions.
2. Query the model using this endpoint and provide data to be used by the model in the request.

import requests
url = 'http://127.0.0.1:47334/api/projects/mindsdb/models/home_rentals_model/predict'

data = [{'number_of_rooms': 2,
  'number_of_bathrooms': 1,
  'sqft': 917,
  'location': 'great',
  'days_on_market': 13,
  'neighborhood': 'berkeley_hills',
  'rental_price': 3901,
  'created_at': '2024-02-24 02:28:21.746167'},
 {'number_of_rooms': 0,
  'number_of_bathrooms': 1,
  'sqft': 194,
  'location': 'great',
  'days_on_market': 10,
  'neighborhood': 'berkeley_hills',
  'rental_price': 2042,
  'created_at': '2024-02-19 06:10:59.693052'},
 {'number_of_rooms': 1,
  'number_of_bathrooms': 1,
  'sqft': 543,
  'location': 'poor',
  'days_on_market': 18,
  'neighborhood': 'westbrae',
  'rental_price': 1871,
  'created_at': '2024-02-12 07:53:45.914146'}]

r = requests.post(url, json={'data': data})
print(r.json())

[
    {
        "__mindsdb_row_id": null,
        "days_on_market": null,
        "location": null,
        "neighborhood": null,
        "number_of_bathrooms": null,
        "number_of_rooms": null,
        "rental_price": 2847,
        "rental_price_anomaly": null,
        "rental_price_confidence": 0.99,
        "rental_price_explain": "{\"predicted_value\": 2847, \"confidence\": 0.99, \"anomaly\": null, \"truth\": null, \"confidence_lower_bound\": 2730, \"confidence_upper_bound\": 2964}",
        "rental_price_max": 2964,
        "rental_price_min": 2730,
        "rental_price_original": null,
        "select_data_query": null,
        "sqft": 823,
        "when_data": null
    }
]

The REST API endpoints can be used with MindsDB running locally at http://127.0.0.1:47334/api.