In the question at:

We are instructed to examine f(x) = exp(x) and fp(x) = Df(f, h=0.001)(x) and graphically estimate their largest difference on [0,3].
The code prescribes the answer "approximately zero". However, when actually plotting out the difference, say with:

f(x) = exp(x)
fp(x) = Df(f, h=0.001)(x)
plot(x -> fp(x)-f(x), 0, 3)

we get something like this figure:

From the figure, the answer suggests itself as .01 not 0.

I don't know how useful it is to put a version number here (1.8) given that the latest number is always changing:

I suggest changing the wording to just "latest stable release" or similar.

serveral should be several. Also, the sentence terminates before telling us what it is that Cubature provides several of.

There must be 12 or so figures that should be redone in Julia. (Newton, extrema, integration, others?)

I refuse to believe that the reasonable answer to this question would be a cable length of 2 meters (6 feet). To span across the Verrazzano Narrows?
Should maybe the limits of integration be -a, a?

The question at:

And the question at:

are identical.

Additionally,

@syms x
limit((x^x-1)/x, x => 0, dir="+")

yields

−∞

which is not an option among the offered answers.

This call to fzero returns the root x=0, while the text is asking for the other root. Initiating at b/2 doesn't get us where we need to go.

Initiating at 260 does give the right answer.

The text talks about not being able to redefine one function type as another, but here we are just redefining an anonymous function to another anonymous function. Should this be instead

g1 = x -> sin(2x^2)

Now, what height of filling will produce half the volume when?

This sentence is really weird. Rewrite somewhat?