leaspy.models.constant_model module

class ConstantModel(name: str, **kwargs)

Bases: GenericModel

ConstantModel is a benchmark model that predicts constant values (no matter what the patient’s ages are).

These constant values depend on the algorithm setting and the patient’s values provided during calibration. It could predict:

last: last value seen during calibration (even if NaN),

last_known: last non NaN value seen during calibration*§,

max: maximum (=worst) value seen during calibration*§,

mean: average of values seen during calibration§.

\* <!> depending on features, the last_known / max value may correspond to different visits.

§ <!> for a given feature, value will be NaN if and only if all values for this feature were NaN.

Parameters

namestr: The model’s name
**kwargs: Hyperparameters for the model. None supported for now.

See also

ConstantPredictionAlgorithm

Attributes

namestr

The model’s name

is_initializedbool

Always True (no true initialization needed for constant model)

featureslist[str]

List of the model features. Unlike most models features will be determined at personalization only (because it does not needed any fit)

dimensionint

Number of features (read-only)

parametersdict

Model has no parameters: empty dictionary. The prediction_type parameter should be defined during personalization. Example:

>>> AlgorithmSettings('constant_prediction', prediction_type='last_known')

Methods

`compute_individual_trajectory`(timepoints, ...)	Compute scores values at the given time-point(s) given a subject's individual parameters.
`get_hyperparameters`(*[, with_features, ...])	Get all model hyperparameters
`hyperparameters_ok`()	Check all model hyperparameters are ok
`initialize`(dataset[, method])	Initialize the model given a dataset and an initialization method.
`load_hyperparameters`(hyperparameters, *[, ...])	Load model hyperparameters from a dict
`load_parameters`(parameters, *[, list_converter])	Instantiate or update the model's parameters.
`save`(path, **kwargs)	Save Leaspy object as json model parameter file.
`validate_compatibility_of_dataset`(dataset)	Raise if the given dataset is not compatible with the current model.

compute_individual_trajectory(timepoints, individual_parameters)

Compute scores values at the given time-point(s) given a subject’s individual parameters.

Parameters

timepointsscalar or array_like[scalar] (list, tuple, numpy.ndarray): Contains the age(s) of the subject.
individual_parametersdict[str, Any]: Contains the individual parameters. Each individual parameter should be a scalar or array_like

Returns

torch.Tensor: Contains the subject’s scores computed at the given age(s) Shape of tensor is (1, n_tpts, n_features)

property dimension: Optional[int]

get_hyperparameters(*, with_features=True, with_properties=True, default=None) → Dict[str, Any]

Get all model hyperparameters

Parameters

with_features, with_propertiesbool (default True): Whether to include features and respectively all _properties (i.e. _dynamic_ hyperparameters) in the returned dictionary
defaultAny: Default value is something is an hyperparameter is missing (should not!)

Returns

dict { hyperparam_namestr -> hyperparam_valueAny }

hyperparameters_ok() → bool

Check all model hyperparameters are ok

Returns

bool

initialize(dataset: Dataset, method: str = None)

Initialize the model given a dataset and an initialization method.

After calling this method is_initialized should be True and model should be ready for use.

Parameters

datasetDataset: The dataset we want to initialize from.
methodstr, optional (default None): A custom method to initialize the model

load_hyperparameters(hyperparameters: Dict[str, Any], *, with_defaults: bool = False) → None

Load model hyperparameters from a dict

Parameters

hyperparametersdict[str, Any]: Contains the model’s hyperparameters
with_defaultsbool (default False): If true, it also resets hyperparameters that are part of the model but not included in hyperparameters to their default value.

Raises

LeaspyModelInputError: if inconsistent hyperparameters

load_parameters(parameters, *, list_converter=<built-in function array>) → None

Instantiate or update the model’s parameters.

Parameters

parametersdict: Contains the model’s parameters.
list_convertercallable: The function to convert list objects.

save(path: str, **kwargs)

Save Leaspy object as json model parameter file.

Default save method: it can be overwritten in child class but should be generic…

Parameters

pathstr: Path to store the model’s parameters.
**kwargs: Keyword arguments for json.dump method.

validate_compatibility_of_dataset(dataset: Dataset)

Raise if the given dataset is not compatible with the current model.

Parameters

datasetDataset: The dataset we want to model.

Raises

LeaspyDataInputError: If and only if data is incompatible with model.