Revise Thresholding epsiode for teaching with Jupyter notebooks

[uschille]

Revision of the Thresholding episode for teaching with Jupyter notebooks. Resolves #34, resolves #36. Includes #154.

There are some pending dependencies on #153 and #157, but I think content-wise this is ready for review.

Currently, images are plotted with the sequence

%matplotlib widget
fig, ax = plt.subplots()
plt.imshow(image)
plt.show()

The call to plt.subplots() ensures that a new figure is created for the current cell (otherwise the image will be shown in a previous output cell in the notebook). With %matplotlib widget, plt.imshow behaves similar to skimage.io.imshow. The plotting/viewing style may have to be changed depending on resolution of #157.

[quist00]

There are some pending dependencies on #153 and #157, but I think content-wise this is ready for review. I think #153 is effectively a dependency for everything we do, so I would be in favor of holding off on any merges until we settle that one and then revise our drafts to be congruent with that outcome.

In the meantime, I will pull your draft local and try to go through it as if I was learning on my own and see how it reads. Looking a the individual commits, they make great sense, but it is just hard for me to visualize how it all fits together without a run through with just the final.

[uschille]

Thanks for the reviews @quist00 and @tobyhodges! I will work on the suggested changes.

[uschille]

Hi @ErinBecker, thanks for joining our co-working session today. I'm adding you to the reviewers for this PR since you brought up #32. I have added some more discussion to the episode on how to choose thresholds - do you think it provides a satisfactory answer?

[uschille]

Great point @K-Meech! I have revised the episode accordingly.