MetaClassifier (starclass.MetaClassifier)

class starclass.MetaClassifier(clfile='meta_classifier.pickle', *args, **kwargs)

Bases: BaseClassifier

The meta-classifier.

clfile

Path to the file where the classifier is saved.

Type:

str

classifier

Actual classifier object.

Type:

Classifier_obj

features_used

List of features used for training.

Type:

list

Code author: James S. Kuszlewicz <kuszlewicz@mps.mpg.de>

Code author: Rasmus Handberg <rasmush@phys.au.dk>

__init__(clfile='meta_classifier.pickle', *args, **kwargs)

Initialize the classifier object.

Parameters:

clfile (str) – Filepath to previously pickled Classifier_obj

build_features_table(features, total=None)

Build table of features.

Parameters:

• features (iterable) – Features to build table from.

• total (int, optional) – Number of features in features. If not provided, the length of features is found using len().

Returns:

Two dimensional float32 ndarray with probabilities from all classifiers.

Return type:

ndarray

Code author: Rasmus Handberg <rasmush@phys.au.dk>

classify(task)

Classify a star from the lightcurve and other features.

Will run the do_classify() method and check some of the output and calculate various performance metrics.

Parameters:

features (dict) – Dictionary of features, including the lightcurve itself.

Returns:

Dictionary of classifications

Return type:

dict

See also

do_classify(), load_star()

Code author: Rasmus Handberg <rasmush@phys.au.dk>

close()

Close the classifier.

do_classify(features)

Classify a single lightcurve.

Parameters:

features (dict) – Dictionary of features.

Returns:

Dictionary of stellar classifications.

Return type:

dict

Raises:

• UntrainedClassifierError – If classifier has not been trained.

• ValueError – If any features are NaN.

Code author: Rasmus Handberg <rasmush@phys.au.dk>

feature_importance_complete(tset=None, features=None, probs=None, diagnostics=None)

Function which will be called when feature importance is finishing.

Parameters:

• tset

• features

• probs

• diagnostics

See also

BaseClassifier.train()

Code author: Rasmus Handberg <rasmush@phys.au.dk>

load(infile)

Loads classifier object.

load_star(task)

Receive a task from the TaskManager, loads the lightcurve and returns derived features.

Parameters:

task (dict) – Task dictionary as returned by TaskManager.get_task().

Returns:

Dictionary with features.

Return type:

dict

See also

TaskManager.get_task()

Code author: Rasmus Handberg <rasmush@phys.au.dk>

parse_labels(labels)

Convert iterator of labels into full numpy array, with only one label per star.

TODO: How do we handle multiple labels better?

save(outfile)

Saves the classifier object with pickle.

test(tset, save=None, feature_importance=False)

Test classifier using training-set, which has been created with a test-fraction.

Parameters:

• tset (TrainingSet) – Training-set to run testing on.

• save (callable, optional) – Function to call for saving test-predictions.

test_complete(tset=None, features=None, probs=None, diagnostics=None)

Function which will be called when training is finishing.

Parameters:

• tset

• features

• probs

• diagnostics

See also

BaseClassifier.train()

Code author: Rasmus Handberg <rasmush@phys.au.dk>

train(tset, savecl=True, overwrite=False)

Train the Meta-classifier.

Parameters:

• tset (TrainingSet) – Training set to train classifier on.

• savecl (bool, optional) – Save the classifier to file?

• overwrite (bool, optional) – Overwrite existing classifer save file.

Code author: James S. Kuszlewicz <kuszlewicz@mps.mpg.de>

Code author: Rasmus Handberg <rasmush@phys.au.dk>

property classifier_model

property random_seed

Random seed used in derived classifiers.

property random_state

Random state (numpy.random.RandomState) corresponding to random_seed.