palantiri.ClassificationPlotHandlers module

class palantiri.ClassificationPlotHandlers.ClassifierPlotHandler(dataset, trained_classifier, **params)

Bases: palantiri.BasePlotHandlers.PlotHandler

Handles all the plots related of the chosen classifier.

build_confusion_matrix(normalize=False)

Building the confusion matrix

:param normalize: if True confusion matrix is normalized.

build_confusion_matrix_figure(figure_layout)

Builds the confusion matrix figure in confusion_matrix_figure.

:param figure_layout: figure layout - plot.ly layout object.

build_prediction_figure(figure_layout)

Building the classifier prediction figure.

:param figure_layout: figure layout - plot.ly Layout object.

build_roc_figure(figure_layout=Layout())

Building the ROC curve figure of the classifier.

:param figure_layout: figure layout - plot.ly layout object.

confusion_matrix()

The confusion matrix.

:return: The confusion matrix as a numpy array.

dataset()

The dataset

:return: The dataset as a dictionary

classmethod from_pandas_dataframe(dataframe, trained_classifier, **params)

Constructing the handler from a pandas dataframe.

:param dataframe: the dataframe form which the handler is constructed. The ‘target’ column should be included in the dataframe.

:param trained_classifier: sklearn classifier (trained / fitted).

:param params: other params.

:return: returns the classifier plot handler object.

n_classes()

The number of classes.

:return: An int representing the number of classes.

plot_confusion_matrix(figure_layout=None)

Plotting the confusion matrix figure with plot.ly’s iplot function.

:param figure_layout: figure layout - plot.ly layout object.

plot_prediction(figure_layout=None)

Plotting the prediction figure with plot.ly’s iplot function.

:param figure_layout: figure layout - plot.ly Layout object.

plot_roc(figure_layout=None)

Plotting the ROC curve figure with plot.ly’s iplot function.

:param figure_layout: figure layout - plot.ly Layout object.

predicted_target_score()

The predicted score - available if classifier has the predict_proba functionality.

:return: The predicted score.

save_confusion_matrix_figure(file_name)

Saving the confusion matrix figure as an html file.

:param file_name: the html file name.

save_prediction_figure(file_name)

Saving the prediction figure as an html file.

:param file_name: the html file name.

save_roc_figure(file_name)

Saving the ROC curve figure as an html file.

:param file_name: the html file name.

trained_classifier()

The trained classifier .

:return: The classifier in the sklearn format.

class palantiri.ClassificationPlotHandlers.ThreeDimensionalClassifierPlotHandler(dataset, trained_classifier, **params)

Bases: palantiri.ClassificationPlotHandlers.ClassifierPlotHandler

Handles all the plots related of the chosen classifier on 3D.

build_prediction_figure(figure_layout=Layout())

Plotting the classifier prediction and saving the figure.

:param figure_layout: figure layout - plot.ly Layout object.

class palantiri.ClassificationPlotHandlers.TwoDimensionalClassifierPlotHandler(dataset, trained_classifier, **params)

Bases: palantiri.ClassificationPlotHandlers.ClassifierPlotHandler

Handles all the plots related of the chosen classifier on 2D.

build_prediction_figure(figure_layout=Layout(), step_size=0.01)

Building the classifier prediction figure.

:param figure_layout: figure layout - plot.ly Layout object.

:param step_size: Plot resolution.