Classifiers

_DEFAULT_CLASSIFIERS

A private dictionary containing the classes of the default classifiers that EasyFit provides in EasyClassifier.

The table contains the keys of the dictionary to access the models, and links to the documentation of each model.

Classifier (Dict Key)	Class (Dict Value)
DummyClassifier	sklearn.dummy.DummyClassifier
LogisticRegression	sklearn.linear_model.LogisticRegression
RidgeClassifier	sklearn.linear_model.RidgeClassifier
RidgeClassifierCV	sklearn.linear_model.RidgeClassifierCV
LinearSVC	sklearn.svm.LinearSVC
SupportVectorClassifier	sklearn.svm.SVC
DecisionTreeClassifier	sklearn.tree.DecisionTreeClassifier
RandomForestClassifier	sklearn.ensemble.RandomForestClassifier
KNeighborsClassifier	sklearn.neighbors.KNeighborsClassifier
MLPClassifier	sklearn.neural_network.MLPClassifier
AdaBoostClassifier	sklearn.ensemble.AdaBoostClassifier
SGDClassifier	sklearn.linear_model.SGDClassifier
GaussianNaiveBayes	sklearn.naive_bayes.GaussianNB
QuadraticDiscriminantAnalysis	sklearn.discriminant_analysis.QuadraticDiscriminantAnalysis

easyfit.classifiers module

class easyfit.classifiers.EasyClassifier(*args, **kwargs)

Bases: _EasyModel

Fit classifier models in

• _DEFAULT_CLASSIFIERS (if include_defaults=True)

• models_dict (if models_dict != None)

Parameters:

• models_dict (Dictionary of additional models) –

  Can hold:

  • classes: models_dict = {‘LinearRegression’: LinearRegression}

  • objects: models_dict = {‘LinearRegression’: LinearRegression()}

  (Default value = None)

• include_defaults (boolean) –

  Include _DEFAULT_CLASSIFIERS in trained models

  (Default value = True)

evaluate(X, y, as_df=True, model_first=True, from_preds=False)

Returns models results on each of the metrics in self._METRICS dictionary

Parameters:

• X (array of features) –

• y (array of targets) –

• as_df (boolean) –

  • if True: return results in pd.DataFrame

  • if False: return results in dictionary

  (Default value = True)

• model_first (boolean) –

  • if True: returns models at axis=0 (rows), results at axis=1 (columns)

  • if False: returns models at axis=1 (columns), results at axis=0 (rows)

  (Default value = True)

• from_preds (boolean) –

  • if True: make preditions then calacuate metrics (X holds input features)

  • if False: calcualte metrics from predictions (X holds predictions)

  (Default value = False)

Returns:

results

Return type:

Dict (as_df=False) or pd.Dataframe (as_df=True)

fit(X, y)

Fit classifiers in self._models on features X with targets y

Calls method fit for each model in self._models

Parameters:

• X (array of features) –

• y (array of targets) –

Return type:

None

get_model(model_key)

Get specific model from self._models

Parameters:

model_key (the key for model in self._models) –

Returns:

• model object corrseponding to key if key exist

• None if key does not exist

predict(X)

Make predictions for features in X

Call predict method for each model in self._models

Parameters:

X (array of features) –

Returns:

preds – Dictionary with same keys in self._models and predictions for each model of features in X

Return type:

Dict

score(X, y, as_df=True, sort=True)

Calculate score for each model in self._models

Calls score method for each model in self._models

Return mean accuracy for each model on the given data and labels

Parameters:

• X (array of features) –

• y (array of targets) –

• as_df (boolean) –

  • if True: return results in pd.DataFrame

  • if False: return results in dictionary

  (Default value = True)

• sort (boolean) –

  • if True: returns results sorted in discending order by score

  • if False: returns results in the original order of models

  (Default value = True)

Returns:

results

Return type:

Dict (as_df=False) or pd.Dataframe (as_df=True)