Please Update the Content of This Video - It's Outdated
First, thank you for all your hard work. I am grateful for this platform and the content you provide.
But please update the content of this video. It's outdated and wrong at this point. I will show you with a screenshot that the behavior that is demonstrated, which is counterintuitive and buggy to begin with, has been fixed by the developers of numpy. And thank goodness it has, because it's confusing. I'm glad they realized it and fixed it.
Next, there is a quiz question that unfortunately references this buggy behavior. I would appreciate that quiz question being reconsidered. It's frustrating to take classes, study for the quizes, and then be faced with bad grades. Now don't get me wrong, if I didn't study and get bad grades, that's my fault. But if we design quiz questions based on buggy behavior that's since been fixed, I would love to see the quizes designed differently.
Thanks for your consideration.
This code makes sense, and is how numpy works now, thankfully. In the course, this throws an error.
But in the course, this code executes correctly. But it should not. It should throw an error as it makes no sense. Thankfully, it now does throw an error like it should.
Hi,
Why would someone ever use np.arange(start=x) without specifying the end point? If it doesn't make sense calling the function with only the start point why would we do that?
Having an old version of some libraries is very common in data science and coding in general, updating libraries after launching the project could lead to massive proplems, I promise you, google, facebook and NASA have projects running right now using this version of numpy, and even more, running on Python 2. That's why the requirements.txt file exists. That's why virtual environments exists. If you want want a code to run like the creator wanted to you need to have the same versions of dependencies.
I have studied datascience in different well-known accredited online platforms and all of them had different versions than I use. Udacity uses the append method in the course material while it's already deprecated and been substituted by concatenate. Why? because versions get updated every second and there's no way to update every course video or excersise with the version that we use right now because for us it takes seconds to update a virsion using pip or anaconda but for them it could take days or weeks for every update to check and update every video and every excersise and eventually it'll lead to a lot of problems too because they won't update the np.arange() function, they'll update the whole numpy library which had updates on other functions that will need to be updated in the course material, excersises and coding environment too. This is what really doesn't make sense and no platform follows.
Understanding how to manage versioning and legacy code is actually a core skill for any Data Scientist. Dealing with these discrepancies now is great preparation for real-world projects where you'll rarely work with the absolute latest versions of every library.
-Justin
