DESCRIBE PREDICTOR Statement¶
The DESCRIBE PREDICTOR statement is used to display the attributes of an existing model.
DESCRIBE PREDICTOR ... FEATURES Statement¶
Description¶
The DESCRIBE PREDICTOR mindsdb.[predictor_name].features statement displays how the model encoded the data before the training process.
Syntax¶
Here is the syntax:
DESCRIBE PREDICTOR mindsdb.[predictor_name].features;
On execution, we get:
+--------------+-------------+--------------+-------------+
| column | type | encoder | role |
+--------------+-------------+--------------+-------------+
| column_name | column_type | encoder_used | column_role |
+--------------+-------------+--------------+-------------+
Where:
| Name | Description |
|---|---|
[predictor_name] |
Name of the model to be described. |
column |
Data columns that were used to create the model. |
type |
Data type of the column. |
encoder |
Encoder type used for the column. |
role |
Role of the column (feature or target). |
Example¶
Let's look at an example using the home_rentals_model model.
DESCRIBE PREDICTOR mindsdb.home_rentals_model.features;
On execution, we get:
+---------------------+-------------+----------------+---------+
| column | type | encoder | role |
+---------------------+-------------+----------------+---------+
| number_of_rooms | categorical | OneHotEncoder | feature |
| number_of_bathrooms | binary | BinaryEncoder | feature |
| sqft | integer | NumericEncoder | feature |
| location | categorical | OneHotEncoder | feature |
| days_on_market | integer | NumericEncoder | feature |
| neighborhood | categorical | OneHotEncoder | feature |
| rental_price | integer | NumericEncoder | target |
+---------------------+-------------+----------------+---------+
Here the rental_price column is the target column to be predicted. As for the feature columns, these are used to train the ML model to predict the value of the rental_price column.
DESCRIBE PREDICTOR ... MODEL Statement¶
Description¶
The DESCRIBE PREDICTOR mindsdb.[predictor_name].model statement displays the performance of the candidate models.
Syntax¶
Here is the syntax:
DESCRIBE PREDICTOR mindsdb.[predictor_name].model;
On execution, we get:
+-----------------+-------------+---------------+-----------+---------------------+
| name | performance | training_time | selected | accuracy_functions |
+-----------------+-------------+---------------+-----------+---------------------+
| candidate_model | performance | training_time | selected | accuracy_functions |
+-----------------+-------------+---------------+-----------+---------------------+
Where:
| Name | Description |
|---|---|
[predictor_name] |
Name of the model to be described. |
name |
Name of the candidate model. |
performance |
Accuracy value from 0 to 1, depending on the type of the model. |
training_time |
Time elapsed for the training of the model. |
selected |
1 for the best performing model and 0 for the rest. |
accuracy_functions |
It stores the r2_score value for regression predictions, the balanced_accuracy_score value for classification predictions, and the bounded_ts_accuracy value for time series predictions. The values vary between 0 and 1, where 1 indicates a perfect predictor, based on results obtained for a held out portion of data. |
Example¶
Let's look at an example using the home_rentals_model model.
DESCRIBE PREDICTOR mindsdb.home_rentals_model.model;
On execution, we get:
+------------+--------------------+----------------------+----------+---------------------+
| name | performance | training_time | selected | accuracy_functions |
+------------+--------------------+----------------------+----------+---------------------+
| Neural | 0.9861694189913056 | 3.1538941860198975 | 0 | ['r2_score'] |
| LightGBM | 0.9991920992432087 | 15.671080827713013 | 1 | ['r2_score'] |
| Regression | 0.9983390488042778 | 0.016761064529418945 | 0 | ['r2_score'] |
+------------+--------------------+----------------------+----------+---------------------+
The value of the accuracy_functions column is stored in the performance column. For example, the r2_score value of the Neural model is 0.9861694189913056.
DESCRIBE PREDICTOR ... ENSEMBLE Statement¶
Description¶
The DESCRIBE PREDICTOR mindsdb.[predictor_name].ensemble statement displays the parameters used to select the best candidate model.
Syntax¶
Here is the syntax:
DESCRIBE PREDICTOR mindsdb.[predictor_name].ensemble;
On execution, we get:
+-----------------+
| ensemble |
+-----------------+
| {JSON} |
+-----------------+
Where:
| Name | Description |
|---|---|
[predictor_name] |
Name of the model to be described. |
ensemble |
Object of the JSON type describing the parameters used to select the best candidate model. |
Example¶
Let's look at an example using the home_rentals_model model.
DESCRIBE PREDICTOR mindsdb.home_rentals_model.ensemble;
On execution, we get:
+----------------------------------------------------------------------+
| ensemble |
+----------------------------------------------------------------------+
| { |
| "encoders": { |
| "rental_price": { |
| "module": "NumericEncoder", |
| "args": { |
| "is_target": "True", |
| "positive_domain": "$statistical_analysis.positive_domain" |
| } |
| }, |
| "number_of_rooms": { |
| "module": "OneHotEncoder", |
| "args": {} |
| }, |
| "number_of_bathrooms": { |
| "module": "BinaryEncoder", |
| "args": {} |
| }, |
| "sqft": { |
| "module": "NumericEncoder", |
| "args": {} |
| }, |
| "location": { |
| "module": "OneHotEncoder", |
| "args": {} |
| }, |
| "days_on_market": { |
| "module": "NumericEncoder", |
| "args": {} |
| }, |
| "neighborhood": { |
| "module": "OneHotEncoder", |
| "args": {} |
| } |
| }, |
| "dtype_dict": { |
| "number_of_rooms": "categorical", |
| "number_of_bathrooms": "binary", |
| "sqft": "integer", |
| "location": "categorical", |
| "days_on_market": "integer", |
| "neighborhood": "categorical", |
| "rental_price": "integer" |
| }, |
| "dependency_dict": {}, |
| "model": { |
| "module": "BestOf", |
| "args": { |
| "submodels": [ |
| { |
| "module": "Neural", |
| "args": { |
| "fit_on_dev": true, |
| "stop_after": "$problem_definition.seconds_per_mixer", |
| "search_hyperparameters": true |
| } |
| }, |
| { |
| "module": "LightGBM", |
| "args": { |
| "stop_after": "$problem_definition.seconds_per_mixer", |
| "fit_on_dev": true |
| } |
| }, |
| { |
| "module": "Regression", |
| "args": { |
| "stop_after": "$problem_definition.seconds_per_mixer" |
| } |
| } |
| ], |
| "args": "$pred_args", |
| "accuracy_functions": "$accuracy_functions", |
| "ts_analysis": null |
| } |
| }, |
| "problem_definition": { |
| "target": "rental_price", |
| "pct_invalid": 2, |
| "unbias_target": true, |
| "seconds_per_mixer": 57024.0, |
| "seconds_per_encoder": null, |
| "expected_additional_time": 8.687719106674194, |
| "time_aim": 259200, |
| "target_weights": null, |
| "positive_domain": false, |
| "timeseries_settings": { |
| "is_timeseries": false, |
| "order_by": null, |
| "window": null, |
| "group_by": null, |
| "use_previous_target": true, |
| "horizon": null, |
| "historical_columns": null, |
| "target_type": "", |
| "allow_incomplete_history": true, |
| "eval_cold_start": true, |
| "interval_periods": [] |
| }, |
| "anomaly_detection": false, |
| "use_default_analysis": true, |
| "ignore_features": [], |
| "fit_on_all": true, |
| "strict_mode": true, |
| "seed_nr": 420 |
| }, |
| "identifiers": {}, |
| "accuracy_functions": [ |
| "r2_score" |
| ] |
|} |
+----------------------------------------------------------------------+
Need More Info?
If you need more information on how to DESCRIBE PREDICTOR or understand the results, feel free to ask us on the community Slack workspace.