def display_digit_and_predictions(image: np.ndarray, label: int, pred: int,
pred_one_hot: np.ndarray) -> None:
"""Display the MNIST image and the predicted class as a title.
Parameters
----------
image : np.ndarray
MNIST image.
label : int
True class number.
pred : int
Predicted class number.
pred_one_hot : np.ndarray
One-hot encoded prediction.
"""
if image.shape == (784, ):
image = image.reshape((28, 28))
_, axs = plt.subplots(1, 2)
pred_one_hot = [[int(round(val * 100.0, 4)) for val in pred_one_hot[0]]]
# Table data
labels = [i for i in range(10)]
axs[0].axis('tight')
axs[0].axis('off')
axs[0].table(cellText=pred_one_hot, colLabels=labels, loc="center")
# Image data
axs[1].imshow(image, cmap=plt.get_cmap('gray_r'))
# General plotting settings
plt.figure_title('Label: %d, Pred: %d' % (label, pred))
plt.show()
def display_digit(image, label=None, pred_label=None):
"""Display the Digit from the image.
If the Label and PredLabel is given, display it too.
"""
if image.shape == (784, ):
image = image.reshape((28, 28))
label = np.argmax(label, axis=0)
if pred_label is None and label is not None:
plt.figure_title('Label: %d' % (label))
elif label is not None:
plt.figure_title('Label: %d, Pred: %d' % (label, pred_label))
plt.imshow(image, cmap=plt.get_cmap('gray_r'))
plt.show()
def display_digit_and_predictions(image, label, pred, pred_one_hot):
if image.shape == (784, ):
image = image.reshape((28, 28))
_, axs = plt.subplots(1, 2)
pred_one_hot = [[int(round(val * 100.0, 4)) for val in pred_one_hot[0]]]
# Table data
labels = [i for i in range(10)]
axs[0].axis('tight')
axs[0].axis('off')
axs[0].table(cellText=pred_one_hot, colLabels=labels, loc="center")
# Image data
axs[1].imshow(image, cmap=plt.get_cmap('gray_r'))
# General plotting settings
plt.figure_title('Label: %d, Pred: %d' % (label, pred))
plt.show()
def display_digit(image: np.ndarray,
label: np.ndarray = None,
pred_label: np.ndarray = None) -> None:
"""Display the MNIST image.
If the *label* and *label* is given, these are also displayed.
Parameters
----------
image : np.ndarray
MNIST image.
label : np.ndarray, optional
One-hot encoded true label, by default None
pred_label : np.ndarray, optional
One-hot encoded prediction, by default None
"""
if image.shape == (784, ):
image = image.reshape((28, 28))
label = np.argmax(label, axis=0)
if pred_label is None and label is not None:
plt.figure_title('Label: %d' % (label))
elif label is not None:
plt.figure_title('Label: %d, Pred: %d' % (label, pred_label))
plt.imshow(image, cmap=plt.get_cmap('gray_r'))
plt.show()