Applied Plotting, Charting & Data Representation in Python

I found in general this course too short and too superficial to become fluent with matplotlib. Module 1 provides philosophical background based on the work of Eduard Tufte and Alberto Cairo, an execellent introduction in the general practices and principles to data visualisation, independent on what tools you use (not python/matplotlib related). Modules 2 and 3 are about the matplotlib architecture, basic plotting (line chart, scatter, barchart, histogram, boxplot) and dynamic plotting (animation and interaction), areas that definitely need to dive a little bit deeper to make the concepts stick. Also module 4 on pandas and seaborn could be more elaborate.

The lectures are based on videos in combination with jupyter notebooks. I am big fan of jupyter notebooks but they combine not very well with video instruction. Too often you see the face of the instructor at times you really want to thinker about the code, sometimes the notebook cells scroll out of the video window so you miss the relation with the initial parts. These factors, together with a fast speaking instructor, needed me to often stop and rewind the video to get full understanding of the material. Not very handy.

For the exercises you will probably need more time than indicated if your main experience where the video lectures and jupyter notebooks from the course. A lot of time (80%) is needed to upfront wrangling the data. This is inline with data analysis practices, however this is a course about visualisation only, so one would expect that the focus of the exercise is more on the visualisation part. Expect to spend a lot of time on stackoverflow and the matplotlib documentation. This time could be reduced if the the course material went a bit deeper in some important areas.

As usual with Coursera courses today you cannot submit your work and get it peer reviewed, neither do peer review of other student's work, if you don't pay for a certificate.

For this topic (plotting with python/matplotlib) I would prefer some good online tutorials, for example:
- Matplotlib tutorial (Nicolas P. Rougier) https://www.labri.fr/perso/nrougier/teaching/matplotlib/
- Scientific Python lectures (J.R.Johansson) lecture 4: matplotlib http://nbviewer.jupyter.org/github/jrjohansson/scientific-python-lectures/blob/master/Lecture-4-Matplotlib.ipynb

I believe these tutorials will bring you faster up to speed with matplotlib than this Coursera course.

This is course 2 in the series. I found course 1 challenging and useful. This was a huge disappointment. Far too much time was spent on the philosophy of visualizations--what makes a visualization interesting/useful. That's good but it should be a minor aspect of the course, not full sections devoted to it. I took the course because I wanted to learn the features of plotting in Python using Matplotlib/Seaborn. Seanborn is barely mentioned. The homeworks are all peer reviewed with the grading criteria so broad it doesn't take much to get full credit. If you attempt the assignment and submit something you will likely get full credit. I didn't learn nearly as much as I hoped and will end up reviewing Udemy's Python for Data Science and Machine Learning Bootcamp for more material on charting in Python.

I expected to have more of a hands-on approach and receive a general introduction to multiple libraries available in Python for plotting and making charts and dashboards. Instead I got a one week intro to theory of visualizations. That would be OK but maybe for a 10 minutes video following recommended literature (the recommended book seems to be a good option). But not for one full week.

2nd week I got an unnecessary in-depth explanation of the backend graphical mechanisms of Python. I just want to plot the thing! To this point I'm not sure if there is something in Python similar to ggplot2 from R, or something like plot.ly that enables the creation of out of the box interactive dashboards. This is the kind of stuff you have to cover, not theory or inner mechanisms of the programming language. I'm dropping from this course and especialization.

Furthermore, peer grading is full of plagiarism from seemingly the same GitHub repositories. Then, if you don't copy/paste and just rename variables and change format, you risk getting a bad grade from some peer reviewers. Assignment #2 is not that hard but it takes time (more than the recommended time). If they gave simpler exercises the first time instead of one "complex" chart, then people shouldn't have to go and copy/paste from GitHub.