Evaluation visualization
Functions to visualize evaluation metrics.
- ecgan.visualization.evaluation.boxplot(data, label, title)[source]
Create a boxplot using mpl.
- Parameters
data (
Union[List[ndarray],object]) -- List of data points or List of Lists containing data points if multiple metrics are tracked.label (
List) -- List of labels for each plot.title (
str) -- Title of the plot (usually the name of the metric).
- Return type
ndarray- Returns
Boxplot image encoded as np.ndarray.
- ecgan.visualization.evaluation.visualize_roc(true_labels, predicted_labels)[source]
Calculate and draw the ROC curve for our model using the predicted labels.
- Parameters
true_labels (
ndarray) -- Ground truth.predicted_labels (
ndarray) -- Labels predicted by model.
- Return type
ndarray- Returns
ROC curve image encoded as np.ndarray.