Using Python to Access Web Data

I just completed the course. It is excellent, but needs a few bug fixes.

The Week 5 Chapter assignment references the wrong exemplar (not wrong, per se, but far more complex than a simpler example covered in the lectures that transfers better to the assignment).

This threw a lot of people and protracted completion of the exercise. A plurality of people who noticed this whined. Apparently, the course shell has not been updated in 2 years.

The code snippets reference the P4E website, which is Dr. Chuck's open source version of the course. The links to the snippets should be given in the course shell, rather than requiring the student to navigate over to the P4E website, dive into the Resources area, then scan to find the snippet.

As one reviewer noticed, there are 2 questions in the Week 6 chapter quiz on REST, but no mention of REST in the lectures, and only a cursory mention of it in Chapter 13 of the text.

I also think that the GeoCoding example covered in the lecture videos needs to be broken down into steps. Perhaps it could be developed incrementally--showing the outputs generated and data structures used at each step. I ended up doing this myself, but I have significant prior knowledge.

There is also little use of functions in the course to-date. It might be good to enforce that as a precept of modular design and to inculcate good habits.

I used https://repl.it throughout the course, despite the recommendation that students invoke Python from the shell and use the platform agnostic Atom editor. Repl is a great learning tool and has a social component these days that has some of the collaboration functionality of forums like Slack. I think the course would run better this way.

Overall, this is one of the better classes on Python out there. I took an introductory Data Science oriented version of a Python course at Harvard and wish Dr. Chuck had been the instructor.

On to Course 4: Using Databases with Python...

jerald_cole @ alumni DOT harvard DOT edu

:-)

Although my specialty is zoology, I signed up and paid for this course (one in a series offered by them) to learn a little about Python which intrigued me for some reason.

The course gives a somewhat disjointed, quick overview of using Python to access web data, but Professor Severance does try to touch shallowly on a good array of the relevant subject matter.

The real sin, however, (and the entire reason I’m taking the time to post this review) is to be found in the course’s abomination of a forum, where the “teaching assistants” are aggressively unhelpful. A teaching forum is intended to be a resource for students who have hit a wall in an assignment and need a few well-placed words of guidance to get them over a hump, but there is none of that found here.

Right around week 5 of this 6 week course was an assignment that a fair number of students had trouble with (judging by the number of questions in the forum) . I read several student to teaching assistant exchanges on the forum where a student just couldn’t figure out that bit of code with the vague “guidance” they had received, and when they had the temerity to persist with their questions, were cut off at the knees by the teaching assistant, and instructed to stop asking . . Really condescending, uncalled for, and shocking, quite frankly, in an educational setting.

I would highly recommend that if you want to learn to use Python to access web data, or any of the other courses in the Python for Everybody series at the University of Michigan, you seek out some other, more professionally run courses where you can get actual assistance if you have a question. There are certainly plenty to choose from.

This third course applies and builds on the basic knowledge from the first two python courses. Especially interesting is the information on regular expressions, which kicks this course off. Following this the course gears up into a higher transmission focusing on using python with web browsers and reading webpages. Beautiful Soup gets covered as well as how to deal with XML data. So far so good.

Unfortunately the course loses itself suddenly in Java Script Object Notation and dealing with Twitter and Google Maps to the point where the amount of knowledge required does not match the content and quality of the classes.

Also ' dr. Chuck' seems more and more concerned with his self-promotion in the bonus video material that is provided each week, rather than focusing only on content. The age of the material (the course was ' shot' in 2015) also starts to show with the assignments that often lack sufficient information for students with English as a second language or jump into a level of complexity that is possible more than the basic student can understand.

The 3 installment of the specialization deals with getting, parsing and reading data from the internet, specifically data in XML and JSON formats.

In comparison with the previous 2 chapters, this one feels quite rushed and overwhelming. Now, I completely understand that it's not possible to explain every single thing in a beginner's course, historically and functionality wise but I couldn't help feeling a bit dumb having to copy-paste some import statements in order to complete the assignments.

There's a lot more to talk about on the topic and I fully agree that the content has to be shrunk to a minimum in order to give a high-level overview of the subject but it could be of great help to the course if there were more assignments, more worked exercises, more lectures that go in further detail.

In the final part of the course, the examples are of reading data from Twitter and Google Maps. In theory, it sounds great and exciting, but in reality, I think it's an overkill for the purpose of the course. There's just too much stuff for a complete noob such as myself. Just a month ago, we were learning what's a dictionary and now we're going through nested dictionaries within lists within dictionaries... quite the jump. Adding to this - explaining what's an API is not done good enough. I only know what APIs do because in work in a SaaS company. If I only had the explanation the course provides, I would be immensely confused.

I'm not trying to say programming is easy, not at all. This particular course though can be done better, so it's easier for students to follow, I'm sure.

As usual, Dr. Chruck is great and it's a pleasure to have him as the tutor here.

I'll continue with the specialization and I hope course 4 will be structured better than this one. :)

Cheers and good luck!

This course have a very interesting and perspective topic - Python mostly is used as a scripting language and there is a lot of web-page parsers around the Web based on Python. My expectations were high. However, the course lectures touches the topic "Access Web Data" very superficial.

Week 2 starts with the regular expressions (week 1 is just a Python installation guide) and I have nothing to say about this weeks material - its useful, I even wrote my first excel file parser using regex after finishing this week.

Week 3 is about sockets and HTTP and how to deal with network in Python. Network is a one big separate topic, I understand that, but some more deep explanation how to use Pythons socket module required.

For example, I still have a very vague idea what this line of code does socket.socket(socket.AF_INET, socket.SOCK_STREAM)), and same things about Carriage Return and Linefeed - not explained, but used in the code examples.

Week 4 is about using urllib and BeatifulSoup to parse HTML

The main question for me was about using as an example pages, that are ready-to-be-parsed and there is a lot of code provided with the course materials. For better learning it is good to rewrite all by yourself. Also big question about Secure Socket Layer - it is in the provided code, but, again, not explained nor in videos or by the book.

Weeks 5 and 6 about parsing XML and JSON, using APIs. And for me Week 2, 5 and 6 were the most useful, but a bit more details about some code that is used in the files provided and more real-world-examples can make this course better.

PS: some code needs to be updated in the course files, but it is more challenging and interesting to do it by yourself.