traveltimes_prediction.models package

Submodules

traveltimes_prediction.models.base_model module

class traveltimes_prediction.models.base_model.BaseModel[source]

Bases: object

dump()[source]

Method for dumping of the model - saving all the necessary data for recreation & usage.

Returns:dict
fit(X, Y)[source]

Method for fitting of the model.

Parameters:
  • X (pd.DataFrame) –
  • Y (pd.DataFrame/pd.Series) –
Returns:

self

static load(model)[source]

Method for recreating the model - creating new instance.

Parameters:model (dict) – dumped model
Returns:BaseModel - instance of class derived from BaseModel
name = None
predict(X)[source]

Method for prediction of the traveltime.

Parameters:X (pd.Dataframe/pd.Series) –
Returns:float

traveltimes_prediction.models.cluster_model module

class traveltimes_prediction.models.cluster_model.ClusterModel(regressor=None, regressor_params=None, clusterizer=None, clusterizer_params=None)[source]

Bases: sklearn.base.BaseEstimator, traveltimes_prediction.models.base_model.BaseModel

Regression model. Perform clustering on training data and train individual regressor for each cluster.

dump()[source]

Method for dumping of the model.

Returns:dict - keys are names of attributes, values are their values
fit(*args, **kwargs)
static load(model)[source]

Method for loading of the dumped model

Parameters:model (dict) –
Returns:obj
name = 'ClusterModel'
predict(*args, **kwargs)

traveltimes_prediction.models.combined_model module

class traveltimes_prediction.models.combined_model.CombinedModel(models=None, models_params=None, stacking_function=<function CombinedModel.<lambda>>)[source]

Bases: sklearn.base.BaseEstimator, traveltimes_prediction.models.base_model.BaseModel

Class - estimator. combination of multiple models.

dump()[source]

Method for dumping of the model.

Returns:dict - keys are names of attributes, values are their values
fit(X, Y)[source]

Method for fitting the model.

Parameters:
  • X (np.ndarray) – matrix of features - SxF
  • Y (np.ndarray) – matrix of outputs - S
Returns:

static load(model)[source]

Method for loading of the dumped model

Parameters:model (dict) –
Returns:obj
name = 'CombinedModel'
predict(*args, **kwargs)

traveltimes_prediction.models.create_model module

traveltimes_prediction.models.create_model.create_model(model_dump)[source]

Method for recreation of the model from dumps.

Parameters:model_dump (dict) –
Returns:
traveltimes_prediction.models.create_model.model_translator(obj)[source]

Function for translating the model name to class & vice versa.

Parameters:obj (class/string) – The class descriptor or attribute .name
Returns:string or class
traveltimes_prediction.models.create_model.params_converter(d)[source]

Fucntion for converting the classes to string representations.

Parameters:d (dict) –
Returns:

traveltimes_prediction.models.time_domain_model module

class traveltimes_prediction.models.time_domain_model.TimeDomainModel(regressor=None, regressor_params=None)[source]

Bases: sklearn.base.BaseEstimator, traveltimes_prediction.models.base_model.BaseModel

Regression model. Split data from each day to X minute windows (according to X[ColumnNames.FEAT_TIME_BIN]). Group windows by time. For each group of windows + adjacent groups to this group train individual regressor.

dump()[source]

Method for for dumping of the model.

Returns:dict - Model dumped as dictionary, with keys as params` names and values as values of the params
fit(X, Y)[source]

Method for fitting of the model.

Parameters:
  • X (pp.DataFrame) – data matrix X for training
  • Y (pd.DataFrame) – data vector Y - outputs
Returns:

self

static load(model)[source]

Method for creation of the model from its dump.

Parameters:model (dict) – Dumped model, keys are params` names, values are values.
Returns:object - instance of TimeDomainModel created from its dump
name = 'TimeDomainModel'
predict(*args, **kwargs)

Module contents