SingleModelContainer¶

SingleModelContainer ¶

SingleModelContainer(
    model_config: BaseMlModelConfig[T, U, V, W],
)

Single model container for machine learning models.

This class provides a container for a single machine learning model. It takes a model configuration and provides methods for training a model and making predictions.

Parameters:

Name	Type	Description	Default
`model_config`	`BaseMlModelConfig`	The model configuration, which includes the learner, predictor, training configuration, and prediction configuration.	required

Examples:

>>> import lightgbm as lgb
>>> from sklearn.datasets import make_regression
>>> from sklearn.metrics import r2_score
>>> from sklearn.model_selection import train_test_split
>>> from factrainer.core import SingleModelContainer
>>> from factrainer.lightgbm import LgbDataset, LgbModelConfig, LgbTrainConfig
>>>
>>> # Load data
>>> X, y = make_regression()
>>> train_X, test_X, train_y, test_y = train_test_split(
...     X, y, test_size=0.2, random_state=1
... )
>>>
>>> # Create datasets
>>> train_dataset = LgbDataset(dataset=lgb.Dataset(train_X, train_y))
>>> val_dataset = LgbDataset(dataset=lgb.Dataset(test_X, test_y))
>>> test_dataset = LgbDataset(dataset=lgb.Dataset(test_X, test_y))
>>>
>>> # Configure model
>>> config = LgbModelConfig.create(
...     train_config=LgbTrainConfig(
...         params={
...             "objective": "regression",
...             "seed": 1,
...             "deterministic": True,
...             "verbose": -1,
...         },
...         callbacks=[lgb.early_stopping(100, verbose=False)],
...     ),
... )
>>>
>>> # Create and train model
>>> model = SingleModelContainer(config)
>>> model.train(train_dataset, val_dataset)
>>>
>>> # Make predictions
>>> y_pred = model.predict(test_dataset)
>>>
>>> # Evaluate predictions
>>> metric = model.evaluate(test_y, y_pred, r2_score)

Attributes¶

raw_model `property` ¶

raw_model: U

Get the trained raw model.

Returns:

Type	Description
`U`	The trained model as a RawModel object.

train_config `property` `writable` ¶

train_config: V

Get the training configuration.

Returns:

Type	Description
`V`	The training configuration.

pred_config `property` `writable` ¶

pred_config: W

Get the prediction configuration.

Returns:

Type	Description
`W`	The prediction configuration.

Functions¶

train ¶

train(
    train_dataset: T, val_dataset: T | None = None
) -> None

Train the model. The trained model can be accessed through the raw_model property.

Parameters:

Name	Type	Description	Default
`train_dataset`	`T`	The training dataset.	required
`val_dataset`	`T \| None`	The validation dataset if needed.	`None`

Returns:

Type	Description
`None`

predict ¶

predict(pred_dataset: T) -> Prediction

Make predictions using the trained model.

Parameters:

Name	Type	Description	Default
`pred_dataset`	`T`	The test dataset.	required

Returns:

Type	Description
`Prediction`	The predictions as a NumPy array.

evaluate ¶

evaluate(
    y_true: Target,
    y_pred: Prediction,
    eval_func: EvalFunc[X],
) -> X

Evaluate the model's predictions against true values.

This method evaluates predictions from a single trained model, typically on a held-out test set or validation set.

Parameters:

Name	Type	Description	Default
`y_true`	`Target`	The true target values as a NumPy array.	required
`y_pred`	`Prediction`	The predicted values as a NumPy array. Must have the same shape as y_true.	required
`eval_func`	`EvalFunc[X]`	The evaluation function that takes (y_true, y_pred) and returns a metric. Common examples include sklearn.metrics functions like r2_score, mae, etc.	required

Returns:

Type	Description
`X`	The evaluation score of type X, as returned by eval_func.

Raises:

Type	Description
`ValueError`	If y_true or y_pred are not NumPy arrays.

SingleModelContainer¶