學生對 密歇根大学 提供的 Applied Plotting, Charting & Data Representation in Python 的評價和反饋

The course can be summarized as: "OK, here are some tools that can be used: now read the documentation, Stack Overflow and some papers that we give you links to". Each week there are just few videos. What I expect from a Course are baby steps and clear guidance about good practices. Of course you can learn it all from the Internet - I am taking a Course to get something I will not easily find elsewhere: a good teacher who will guide me through optimal approaches.

I don't think the tutorials walk us through what we are supposed to do. I find myself on youtube watching free tutorials on the very subjects I am paying to learn.

The grading by peers system coupled with the unlocking next week lesson is really aweful. I can't change to previous session event though I worked through the first lesson fast. Everyone has the same issue on the forum but nobody dares to reply. We get no help nor assitance from the platform. This course is purely money stealing. Run away from this course as soon as possible.

This is an interesting course, but the professor really does not spend enough time teaching the topic. It's like as if he expects that giving us a very high level overview on a subject (e.g., "These are the principles of beauty to follow when making a chart!"), followed by 1-2 very specific cases ("here's how to build a scatterplot!") is enough. We're then expected to teach ourselves in order to be able to turn in assignments. I understand that a core skill of any programmer is the capacity to search for code snippets online as well as ask questions to the community, but for an introductory course on Matplotlib, I'd expect more teaching of the subject matter.

The lectures are overloaded with too much information, and the concepts are presented in a complicated way. I wish the courses in this specialization were self sufficient. It just does not feel like I am getting proficient at any of this even though I can get the assignments done on my own. there should be ample practice exercises with the aim of burning the syntaxes and concepts to memory which is usually not the case. This eventually leads to half hearted learning where students are expected to do every thing on there own. I am disappointed. The course can be improved by going deep into the concepts and providing additional resources for students to explore.

Not very good instructions. The Assignments required just self learning online via other video tutorials/documentation....which defeats the purpose of taking an online course

Not a great course. Prof is obviously smart, but the lectures breeze through the material far too quickly and too lightly, with students left to do most of the work themselves via the assignments. I'm a fan of learning by doing, but I question the value of a course when most of what I pick up I get from stack overflow. The assignments aren't well explained or maintained, and the same questions keep coming up from students year after year.

Prof would be well advised to revisit this course, expand and update the content, and clarify the various points of confusion in the assignments.

Of the 6 Coursera courses I have done to date, this was by far the most tedious and frustrating.

There are a few different approaches to creating images in Python using Matplotlib, and this course didn't manage to set any of them out in a cohesive way that was easy to understand or implement.

An intent of the course was to educate learners about the more detailed Matplotlib control, features, e.g. canvases, so they had more control. But the course presentation is so incohesive that learners are just left utterly confused when it comes to doing the assignments, and they end up trying to pull together a mishmash of code from the internet to try and provide a credible assignment response. It is just such an inefficient use of the learner's time.

This course needs to be torn down, the assignments reviewed, and then the lecture material rebuilt in a way that will enable learners to easily implement the points from the lectures, and eliminate the chasms between course content and what's needed to do the assignments.

The instructor just read a text without any interest and passion.

This course was not nearly as valuable to me as the first course in the series. It breezed through a bunch of different plot types without explaining in enough detail what they would be used for or when you should choose to use them. At the same time, it also didn't provide enough clear examples of how to do basic things in matplotlib, which seems to me to be a very non-intuitive thing with poor documentation. I found the first assignment to be very difficult.

Peer review of assignment is very time sensitive. I don't feel it useful.

assignments are unclear and provide few explicit resources for users less familiar with statistics. crucial topics are not discussed and we are instead told to go google them for ourselves, which i could have done without paying for a course.

CONTENT: The instructor shows some examples of different plots in python (e.g. line, bar, scatter) and some concepts (e.g. histograms or heat maps) but doesn't properly explain anything. Mostly you'll get an example graph with snippets of code only working for that particular example and for the assignment you're "strongly encouraged to use other sources". That's not what you're supposed to get when you're paying for an online course. You should get proper explanations.

ASSIGNMENTS: You're basically told to get data from any source you like and then plot some graphs. If you've had some experience with python and got your explanations for plotting from somewhere else, you'll mostly spend more time looking for data to present than for the actual assignment.

I don't understand why there's no selection of graphs and data sets to choose from so you can concentrate on programming and properly presenting data rather than wasting your time looking at reddit like recommended by the instructor.

ASSIGNMENT GRADING: You’ll have to grade your peers’ assignments with a rubric that’s just not working: you can give points for someone uploading an image/writing a paragraph of text, but you have to either give 0 or 100%, so there’s not way to properly grade partially wrong answers. Example: yes, there is an uploaded image and the student has explained how it follows “Cairo’s principle of beauty”, but it doesn’t follow the principle of beauty. So, how to grade: zero or hundred percent?

Likewise, your assignments are graded by your peers, so you’ll usually have at least one or two days to add to each assignment. You should take this into account when opting for the monthly subscription. Additionally, neither you nor your peers are qualified to grade the assignments, because you’re just learning how to curate and present data (if you’re not already a scientist and just want to learn how to do this in Python).

DISCUSSION FORUMS: You won’t find answers or discussions in the discussion forum. There are only posts asking to please grade a student’s assignment because it is urgent because the subscription is ending soon (see above).

SUMMARY: If you need the certificate for Applied Data Science in Python, you probably must take this course. Otherwise I strongly encourage you to skip it and find other (better) resources to learn plotting in Python.

Pros: Very nice assignments

Cons：Instruction is terrible (week 1 on how to draw neat chat is great though). However, coming to the charting, it is terrible. At the end of the course, the instructor didn't emphasize differences among 3 different charting methods:

stackflow has a great post on this: https://stackoverflow.com/questions/37970424/what-is-the-difference-between-drawing-plots-using-plot-axes-or-figure-in-matpl

The instructor should have talked about this in a laymen language in the very beginning.

Although I learned a lot, but mainly from the assignment and stackflow, not week2-4 education videos.

The course is not balanced - lectors give very brief explanations and doing it very fast. There is not much little connections to practical applications. The assignments often are vague, many times I had to research what they actually wanted instead of actually stadying Python. I finished it, but this course creates more frustration than dophamine. Not recommend it

Cant submit for two weeks but billed monthly, this is bogus.

The course teaches you about important plotting and design skills but the peer reviews and lack of clarity for the assignments makes this class less useful than it could be. In the previous class the discussion forums were an invaluable resource but in this class they are full of spam requests for a peer reviews which makes finding useful information more difficult than it should be.

The course takes into account the theoretical approach when creating charts which is something I have never thought of! And I don't think you'll find instructors that will go that deep into theory instead of programming.

To be honest, I don't believe that charting needs any programming skills at all, it is similar to creating front-end apps, so I think their focus (a week is dedicated for it) on the theory was a great choice.

Loved the course! This course teaches you details about matplotlib and enables you to produce beautiful and accurate graphs.. Assignments are challanging, and helps to build a solid foundation.

Course materials are more than 4 years old. That describes not only the videos but also the age of the various Python libraries that you will be working with. The charting and data visualizations discussed are the really the most basic kind that you might find more easily executed in Excel or Google Sheets. My idea of data science visualizations are not bar charts, scatters, line graphs, and box charts.

The age of the course also means that there will be almost no discussion of more contemporary data viz libraries like Seaborn or Plotly or ones I may have never heard of as this course sought to cover the most basic elements of matplotlib. Nothing presented I would say rose to the current professional and academic standards of data visualization.

Very unresponsible professor. No passion at all!!!!! Did not explain the fundamental concept well. As a result, I do not think I have a deep understangin at all. I spend most my time google in order to finish my assignments.

I give him the second star only because I the way he designed those challenging assignments. He should include some skill needed to finish the assignments.

Didn't learn anything, have been paying for over 2 months, with very poor instructions and very tough quizzes, designed to make you search and search without understanding the fundamentals. The outcome? You paying, without really have any fundamentals rooted. Decided to cancel this specialization after various attempts to push through, without feeling that I am getting the value for what I am paying for. I am better off buying a $20 book from Amazon on Pandas and Visualizations and researching each exercise and apply my learnings on Kaggle datasets. This is unfair!

This course makes you give up on data science and MOOCs.

Seriously, the content is poorly presented he keeps on speaking , telling 2-3 lines about a function and so on.

I highly recommend stay away from this pathetic specialization.

No real teaching. Just skimming fast through a table of contents. It's better to read a book on the subject than taking the course.

Good course to learned matplotlib and other Graphs libraries, but the course goes further than Python and also encourages the studies to create more meaningful and beautiful Graphic views.