I create a matplotlib figure within my neural network training loop for creating a Tensorboard image. Therefore, I use plot_to_image() to convert the figure to an image as mentioned in the official Tensorboard guide.
def plot_to_image(figure):
"""Converts the matplotlib plot specified by 'figure' to a PNG image and
returns it. The supplied figure is closed and inaccessible after this call."""
# Save the plot to a PNG in memory.
buf = io.BytesIO()
plt.savefig(buf, format='png')
# Closing the figure prevents it from being displayed directly inside the notebook.
plt.close(figure)
buf.seek(0)
# Convert PNG buffer to TF image
image = tf.image.decode_png(buf.getvalue(), channels=3)
# Add the batch dimension
image = tf.expand_dims(image, 0)
return image
for epoch in range(n_epochs):
# ...
# error_fig = >some fancy plt.fig for visualizing stuff in Tensorboard<
tf.summary.image(name='train_error_img', data=plot_to_image(error_fig), step=epoch)
print('Some metrics for this particular epoch..')
# ...
The plt.close(figure) should prevent the figure to be shown. However, in my jupyter notebook, I get an empty space in my output between the print statements. If I highlight the output, I can even see the three images I create for each epoch as a blank space (Yes, I call the function three different times for one epoch for different figures).
My question now is:
How can I change this behaviour but still get my print statement shown?
Thanks in advance
