It was hard and I could not have passed nor learned much except frustration if it had not bee for Sophie Green (TA) superb support.

I could not start the course on time but the first week was easy. So I was surprised by the work excepctation from the 2nd week. Also it did not match what was forcasted by instructors. 2h --> 10h, 4h--> 20h... and I'm not count the night thinking of ways to solve the problem.

I think that difficult comes from the expectation that documentation is understandable by newbies.

Also question were often tested on type but the expected type output was not mentioned in the questions.

Finaly, I think personally would need to learn how to debug properly a python program (going step by step in it)

Hard and challenging. Thank you Sophie

This is a good introduction to Python and especially pandas for handling data. However, the course material is not very comprehensive, and you are expected to read online documentation and search StackOverflow to find answers to most of the required functionalities if you wish to finish the assignments. But, and here comes the but, this is actually how you would proceed the day you are faced with a "real" data science task, so from that perspective it is a good lesson. Also, for learners out there not wishing to pay for similar material, there are plenty of notebooks (on e.g. GitHub) that more or less contain the same (for free!). In summary, mr. Brooks does a good job explaining the material and the assignments are hands-on and well thought through.

Good introduction to Python, with a heavy focus on Pandas. Definitely worth doing if you're struggling a bit with the Pandas documentation. The course assessment were a bit of an up-hill battle for me, but I feel more skilled for completing them, so I would encourage others to fully engage as much as possible. Same with the readings that are set. In fact, I'd like to see more recommended readings, along the lines of David Donoho's paper. The course uses Jupyter notebooks for assessments, which was refreshing, and has in-video code to work-through which was also much appreciated. All-in-all, take the course if you're interested in Python and Pandas. It will eat your time quite a bit to do the assessments if you're like me, so be prepared for that.

This course was great to get a start in Data Science. I thought the lectures and the course notes were very well presented and served as a good resource for the assignments. The use of Jupyter notebooks was especially helpful. I found the course forums also very helpful in debugging my code. I have some previous experience with Pythons so the level of the assignments was easy enough that I didn't feel overwhelmed, but challenging enough for me do a lot of independent study.

A criticism I have is that the Jupyter notebooks kept logging me off so saving assignments regularly failed and I had to redo work often.

I usually don't pay for courses on Coursera, but the quality of this was so good that I will definitely continue with the specialization.

The course is overall very good and is definitely a very complete introduction to the topic.

Sometimes I found the pace of the course a bit too fast: it's quite difficult to focus on the Professor's explanations and, at the very same time, follow the commands typed on the jupyter notebook (not to mention reproducing them on my own!!). The python exercises are overall quite well done, even though in some parts they could benefit from clearer explanations (thankfully almost always provided in the discussion section). Be careful that often the programming assignments go beyond what has been explained in the lectures and therefore require an active search in the documentation, on online forum such as stack overflow or in the resources section.

The video-based training for the course was good. I think there is often too big a jump from what is covered in the videos to what the learner is asked to work through in an assignment. Perhaps more, and more modular assignments would help. It would be impossible (at least in my opinion) to pass the course without the teaching assistants and the interactions in the forums. (A big hand for the TA's!) The additional readings and thought-provoking questions are very good. Please continue them. Finally, whatever you can do to get Coursera to improve the (incredibly cryptic) automated grader functionality would be appreciated: there were times when I felt that the only way to get an answer right was to sacrifice a goat at the grader's altar.

This course gave me a workout! I had taken previously taken the first two units of Py4e and I should have taken more introductory courses before this one. Videos introduce a lot of material but go too fast and sometimes are hard to follow. Most learning occurs using other resources - Help manuals, StackOverflow, web videos (I found Data School easy to understand), etc. A lot of self-directed learning that will help me progress and learn independently in the future. The tutors were very helpful most of the time - sometimes responses weren't sufficient for my limited background & I needed followup questions. Overall, very satisfied and feel like I have achieved a good introduction into pandas, but still lots of learning needed.

I would say this course is kind of a crash course on pandas. If you are looking to learn pandas from basics and want to explore each possibility of different methods and attributes of pandas object and classes, this is not for you. But, if you already know a bit of pandas and want to build on that, I definitely recommend this course. The course instructor speaks very fast and sometimes it's kinda hard to follow along with video and notebook exercise. The assignments are a bit challenging and requires double the time mentioned in the course if you are not fluent with pandas. I would look to see some built in interactive practice tool within the course rather than separate notebook files. I wish you good luck with the course.

This was more challenging than expected, worth it though. Had to self learn a lot. Previously had some exposure to the basics of Data Science and still found it hard. I would recommend this if you are willing to put some time into it. I would suggest the course instructors to add more lectures and provide more specifics and nuances/methods in lectures. I would also recommend interested learners to enroll in UCSanDiego's Data Science Program. I felt they had more comprehensive lectures. Though at times there were some quirks that I only learned from the current course. Also just a tip, TA's are great and most of the questions you may have might already be answered in the discussion forum, so do check it when you get stuck.

This course was very beneficial. It provided the entire spectrum of difficulty-from easy to quite difficult. I came into this course with a beginner to moderate level of programming knowledge and it proved to still be a mental battle at points. I would say a bit of advise would be to help clarify some of the assignment questions. There were several questions where I wasted a lot of time just because the prompt was not clear. I was able to figure out by asking or by looking through the discussion forums, though. But I think there may be some clarity needed to reduce the time of doing this going forward. I look forward to taking the next courses as well. Keep up the good work and continue to improve the course!

I thought the videos for this course were appropriately concise and well-done. The projects were about the right difficulty and length (although the one for week three was more time consuming than expected). They were interesting enough to keep engaged. I think the span of topics was also useful and appropriate. I would have liked if the autograder gave more detailed output on what was wrong with a submission. I also think it would be extremely beneficial to see the "right" ways of solving this problems after a submission is completed. I'm sure I didn't use the optimal pandas approach in some instances, as it stands I seem to have no way of knowing and improving my knowledge beyond getting a passing grade.

I like the content covered in the lectures. That's very useful for me to clean data parsed from webpages. There is one thing I hope the course can improved, which probably has been mentioned in the forum for many times: the auto-grader is not flexible that it doesn't grant any points to answers that is slightly different from the submitted form (e.g. upper/lower cases, white space, etc.) and we have a hard time figuring out what exactly the auto-grader is accepting. I would suggest modifying the auto-grader such that it allows flexibility in answer acceptance, or breaking down questions into smaller parts (more partial credits) and give more precise/detailed description on the format of answer expected.

There were some small draw-backs, like the fact that the subtitles in the lectures often had mistakes (it is clear it was either computerized or outsourced). But overall the content and the lectures were of very high quality. So I would recomment it to colleagues for sure.

However, there were some very frustrating omissions in the assignments that made them take much longer than needed. Lack of specificity in requirements, or even just not including very crucial details. I had to search the forum to find out that I was doing everything right, but just was doing not exactly the calculation desired. This is a big oversight and cost me time on the assignments that had nothing to do with my understanding.

In general, the course is present in a coherent and concise way. This course mainly focus on pandas library in python. Finishing this course, you will have the essential tool to manipulate pandas series and dataframe. However, the course has room for improvement, especially the assignment section. Question 2 of week 3 assignment is ambiguous and the data set provided is prone to misleading. And I also think the data set provided for week 4 assignment is corrupted. When set the house price data's index to ['State', 'RegionName'], the multi index is not unique! Also, few of the university town names are not in the house price list!

In a nutshell, if you are new to pandas, this is a good place to start.

The homework assignment really require a lot of research, which is daunting. It's not as if the things are that hard to research, but maybe sometimes some hints of where to start looking would be nice. I paid for many more months than one to complete this course, some of which was my unavailability, some was the quantity of time it took to churn through the assignments. OTOH, I had only coded in Netlogo and R, and it took a REALLY long time to understand slicing in python. It's not intuitive and it's not explained well online. How about some slicing questions for practice at the beginning, or separately from the practice? Or a practice module that offers endless slicing practice?

The course is informative and well structured. The programming assignments are challenging, but I consider this a good thing as it forces the students to learn on their own. Be prepared to spend more time than advertised.

The course aims to be realistic and thus puts a lot of emphasis on data preparation and cleaning. However, it lacks tests for your code so it is often tough to figure out exactly what the autograder wants to see prior to submitting. This had caused me some frustration as the final project does not expect a thoroughly cleaned data set, but rather one that is cleaned up to a certain point (e.g., you have to remove parenthesis but leave orphan punctuation in place).

I liked the assignments, but I would've loved to see the instructor's answers at the end of each one, since in many cases you could work around to get to the answer in a non-pandorable way (or what you might thing was pandorable, but wasn't really). I know being an online course it is very difficult to post the answers because you can't control where the answers may end up and how people would use them, but seeing some pandorable answers is really what I wanted to learn in this course (understanding the pandorable way to do it). This worked well in the in-video assignments, where you get to an answer and then you were shown the instructor's answer, which sometimes surprised you.

I definitely learned a lot of Python skills from this course, and it was challenging (most of the time, in a good way).

However, there were several times where I felt the assignments could have been more clear with expectations. The autograder also made it very difficult to troubleshoot wrong answers without spending time looking through forums. The use of an outdated version of Pandas in the autograder was also a nuisance.

You will definitely need to spend time on Stack Overflow for some assignment questions - but I suppose that's just part of programming. I still do think the course videos and sample problems could have used a few more examples that relate to the assignments.

The course was amazing but it seems to have become outdated. Firstly, some of the codes that worked perfectly on my personal Jupyter did not function correctly on Coursera's Jupyter portal. Also, some discussions in the forum gave some answer hints that were outdated. On Week 4's assignment last question, i.e. the t-test one, I wasted hours trying to get the pvalue 0.0027xxx which was hinted in the discussion forums by a teaching assistant. However, once I submitted my assignment with a different pvalue, I realized my answer had been correct all this time.

Nevertheless, above doesn't undermine the fact that the lectures were amazing and coursework was designed appropriately.

Overall, this was a good introductory course to data science with python. With some previous python experience, I was able to follow along with most examples and assignments. I had not worked with Jupyter notebooks previously, so this course served to get me comfortable with using both notebooks and the pandas library. I found the reading assignments on [data] science to be enjoyable, and perspective broadening. In my opinion, the course somewhat underestimates the time required for assignments by someone not familiar with pd (nontheless the assignments themselves were useful). The existing message boards/forums were useful for figuring out problematic assignments.

This was a good introduction to Pandas. I would recommend knowing the basics of Python before trying -- I had never used Python before and found it challenging at first. The only thing I didn't like was that some of the assignment instructions were ambiguous, and it took longer to get the answers in a format that would be accepted by the autograder than it did to complete the assigned tasks. So I felt that wasted a lot of my time. However, the instructors were very good about answering questions in the discussion forums, so when I got really stuck I could get help. I learned a lot of very useful skills, especially for data cleaning. Thank you!

In general, this course is able to introduce the basics of data manipulation in pandas. Exercises are designed in a way that key python functions can be applied. The downside is, I feel like the grading some what less intuitive. For example, in may assignments, student has to import pandas or numpy in each and every function in order to get the score. Another problem when submitting your assignment is that students have to SAVE changes in order for the new code to take effect. Last but not least, the reading of the grading report is unclear. Students may not get the full reason for certain errors, thus creating challenges to improve their homework.

It was a very good experience and I liked so much having the opportunity to apply what I was learning in the course in every assignment. The only thing I think is not good enough, is that some of the assignments have results that are difficult to find. You are doing a very good code, it works fine, but then you realize that it isn't what they were looking for because the question wasn't clear. Also, at the quizzes, I found it a little annoying when they asked me for things that weren't explained, like specific functions. I could just search them on google and find out how they work, but I'm not sure whether it is the objective of this course.

The assignments are at a perfect introductory, yet not too trivial level.

However, I think the way the assignments are graded is very restrictive (i.e. the auto-grader expects very specific outputs with properties which are not essential to the solution required) and a lot of the student's time is wasted on rewriting the solution to accomodate this.

The "statistics" part of the course is unnecessary, as the only way in which it is relevant is one line of code in the fourth assignment. People who want to know what running a t-test means should be referred to Wikipedia.

Overall I had a great experience and I will do more courses on Coursera.

Personally, I find that the course is good. As it's name, the course teaches how to use Python in many basic Data science task.

I really appreciate the assignments, they cover well the introduced concepts.

However, the only negative point (for which I do not give the course 5 stars) is the speaking speed of speakers. I bet that there are not many students whose IQ is over 140 to understand the videos without pausing several times. I mean, the instructor should focus on the ability of listeners, there are so many times I even could not read what are being appeared on the screen while the instructor are keep speaking and speaking.