Thanks for your suggestions. In the case of the ones that are basically stylistic (e.g. always formulating code as functions), I don’t think there’s much chance that the course developers will consider that worthy of their time. But I will file an enhancement request and point them to this thread.
For the question about how the reshape is done, have you seen this thread? I think I may already have sent that to you in response to an earlier thread you wrote.To get the full value there, make sure to read all the way through to learn about “F” versus “C” order.
It’s entirely possible I’m missing your point here, but I don’t see how floor is relevant in predict. What is floor(0.75)? Perhaps you meant round(), but then you have to run the experiment to figure out how it handles the value of exactly 0.5. Or we can just write exactly the code we want, meaning using > 0.5 as the criterion for “Yes”. If I were going to suggest an enhancement here, it would be not to use a for loop and to use “logical indexing” instead, which is a more elegant way to express the same thing. Try running the following code and watch what happens:
A = np.random.randint(0, 10, (4,3))
print(f"A before =\n{A}")
A[A > 7] = 42
print(f"A after =\n{A}")