The find() Method
Description
The find() method is used to get predictions from the model table. The data is
not persistent - it is returned on the fly as a result-document.
Syntax
Here is the syntax:
db.predictor_name.find({column: "value", column: "value"});
On execution, we get:
{
"column_name1" : "value",
"column_name2": "value",
...columns
"select_data_query": null,
"when_data": null,
"target_name_original": "value",
"target_name_confidence": "value",
"target_name_explain": "{\"predicted_value\": value, \"confidence\": value, \"anomaly\": null, \"truth\": null, \"confidence_lower_bound\": value \"confidence_upper_bound\": value}",
"target_name_anomaly": "value",
"target_name_min": "value",
"target_name_max": "value"
}
Where:
| Expressions | Description |
|---|
"target_name_original" | The real value of the target variable from the collection. |
"target_name_confidence" | Model confidence. |
"target_name_explain" | JSON object that contains additional information, such as predicted_value, confidence, anomaly, truth, confidence_lower_bound, confidence_upper_bound. |
"target_name_anomaly" | Model anomaly. |
"target_name_min" | Lower bound value. |
"target_name_max" | Upper bound value. |
Example
Making a Single Prediction
The following MQL statement fetches the predicted value of the rental_price
column from the home_rentals_model model. The predicted value is the rental
price of a property with attributes listed as a parameter to the find()
method.
db.home_rentals_model.find({sqft: "823", location: "good", neighborhood: "downtown", days_on_market: "10"});
On execution, we get:
{
"sqft": 823,
"location": "good",
"neighborhood": "downtown",
"days_on_market": 10,
"number_of_rooms": null,
"number_of_bathrooms": null,
"initial_price": null,
"rental_price": 1431.323795180614,
"select_data_query": null,
"when_data": null,
"rental_price_original": null,
"rental_price_confidence": 0.99,
"rental_price_explain": "{\"predicted_value\": 1431.323795180614, \"confidence\": 0.99, \"anomaly\": null, \"truth\": null, \"confidence_lower_bound\": 1379.4387560440227, \"confidence_upper_bound\": 1483.2088343172054}",
"rental_price_anomaly": null,
"rental_price_min": 1379.4387560440227,
"rental_price_max": 1483.2088343172054
}
Making Bulk Predictions
Bulk Predictions WIP
The bulk predictions is a work in progress.
