This course is mostly great. The automatic grading software usually provided helpful output, and when it didn't there were discussion threads with staff-posted answers that unblocked me, so I was never stuck. The assignments were great because they are generally real-world scenarios. There was a slight misalignment between the week 2 teachings and assignment which required some reading ahead and using StackOverflow, but otherwise the assignments aligned well with their weeks' teachings. The videos are easy to understand. I feel super comfortable with python, pandas, and Jupyter notebooks after taking this course.

The Videos were usually engaging, although the pacing was often too fast, and the video did not always match the words spoken, this was very distracting. The exercises and readings were excellent as they focused on big picture issues and the ethics of data science, which kept the course interesting as data science and coding can tend to be quite dry. The assignments were my favourite part as they were stimulating and often challenging. They really forced you to go back to the course content, and to the pandas documentation to figure things out. All in all a really enjoyable course, which I learnt a lot out of.

It is a really great course that forces you to rely on your ability to teach yourself, but has a great discussion forum that all students should be on to get the full understanding. I would like more feedback from the autograder and I would appreciate a bit more of a warning about the time estimate per week. Definitely 8-10 hours a week just in assignments. This is mostly due to the fact that the autograder can get you stuck, also make sure to use pandorable code because if you don't most likely your answer won't be correct. Overall learned a ton and can't wait to start the next course in this specialization!

This course is much more time-consuming than it claims. As a person usually working with R and who has taken a beginner specialization of Python, I do not think the concepts in this course are difficult. However, the design of assignments makes you spend substantial time on tedious tasks like fixing some minor issues to make answers exactly the same as it is designed. An example is the order of entry is required to be the same as the pre-designed answer. Debugging these unimportant errors (by unimportant, I mean they will not affect outputs in a practical way) makes people really struggling and desperate.

To course content creators:

Thank you for working on keeping the course materials updated. While this is really important, it is really frustrating that a course is upgraded and you are forced to switch mid way through. I'm at the end of week 3 and as a new mom I have very limited time to work on my course work. For full time working learners/single parents/learners with several other commitments, the flexibility offered is of most importance. It's inexcusable to ask learners to complete a course in limited time when everyone signed up and paid for a course that you can take your time to complete.

I had to look up additional information for Assignment 4 and, to my surprise, came across a completed version. The autograder did not appear to work appropriately for Assignment 4 and, after trying for hours and not understanding why the autograder did not recognize the type(...) of my answer, I ended up copy-pasting someone else's response to Assignment 4, Question 1) to achieve the 80% threshold; this ultimately allowed me complete the assignment.

I apologize deeply for this, but I still have no idea why the autograder was not recognizing the type of my submission (i.e., np.float).

Regardless,

I think the structure of the lessons are highly relevant for data sciecne and taught well. So for Syllabus content I think you are doing veryy well

Cons: Nearly all the problems I encountered were related to the autograder not being helpful! There could be better error handling/debugging i.e. instead of just saying the answer is wrong, explain why and the aswer that is expected. For example, I spent DAYS on the final question (run ttest)of the last assignment. I had the right answer all along except the autograder was expecting the type to be numpy float 64, whereas my answer was just a float!

Course content is great and so the way how videos are made. Issue (at least from my point of view) is the test and evaluation part. I love challenges and the 1st test I really enjoyed, but for the next 3 tests it was not possible to finish whole test without reading additional information in disscusion section for each week. I consider it as annoying. It would be great to add some extra informations into the description of questions. Evaluation is kinda problematic since sometimes you manage to get the required output but it is still considered as wrong. Other than that very nice, thanks :).

What I like this course:

Assignment! Geat assignments. Challenging and learnt a lot.

What I do not like this course:

instruction videos: way too shallow, and no real examples. I understand pandas and numpy are both very comprehensive libraries and it is impossible to cover every details. However, if the instructor could teach a real case study,(like similar cases in assignments), learners could have better understading going into the assignments.

Finally, teacing staff are great, very helpful.

I learned a lot from this course, not from the instructors, but from assignments and teaching staff.

I can see a couple of problems in this course. How to access and change elements in a data frame (especially the use of loc and iloc) is not explained very well and I believe there is an error in the slides. The second thing is that I have found more powerful methods in pandas documentation, for example assign() or nlargest() and nsmallest(). This is not really a problem if one of the "hidden" purposes is to push students to actively search the docs for what they would like to do.

Work load for the assignments is just fine in my opinion, not too little, not too much.

Overall I am satisfied.

I felt that there was very little hand-holding in this course. I think there was good and bad from that. On the one hand, I think it could have been more helpful if the lecture content more directly correlated to the assignment questions. However, I definitely think that learning to utilize the internet (stack overflow etc.) is an integral part to coding in any language in today's landscape as well. So there is something to be said for that for sure. Overall, I think I learned a lot of valuable skills in the course and I'm looking forward to continuing the others in the specialization.

I enrolled 4 years back this course but that time got some issues as some video lectures were not much clear or comprehensive (not the lectures from Christopher Brooks). Besides , submitting assignments were much easier as I got required assistance from staff , specially I want to thank Yusuf Ertas. I wanted to give 5 starts but given 4 stars as some of the questions/problems in assignments were really confusing to understand where the actual task/question was quite simple to solve . This issue frustrated me badly and costed me to extend course accomplishment time twice.

It is a very nice course, providing you are able to pour a LOT of individual efforts into it, especially into the assignments. The knowledge required to finish the course is far beyond what they taught. I spent 70% of my time in the course on Stackoverflow or Google, or asking my friends. However, I admitted I gained huge amount of knowledge and experiences, as well as "individual research skills" and confidence, from this course. I would recommend going through this course in a much shorter period of time than recommended (like in a week), during your vacation or week off.

Solid introduction to using Python for basic data science applications. By the end of the course I feel confident that I know how to go about cleaning data and doing some basic analysis on it. The practical programming assignments are great, you can really dig your teeth into them without them being so challenging that it scares you off. Expect it to take you longer to complete the assignments than the official estimated time.

What's also great is that all the code from the lectures is posted so you can easily play with it to be able to follow along with the lectures.

I didn't utilize the discussion pages until the third week, but they are extremely helpful. The pseudo code that was sometimes given in those discussions was also helpful for learning the material because it revealed programming solutions that I was unaware of given my inexperience with the python language. My only real issue is that I wasn't expecting the self-learning for the end of week assignments, but it wasn't that bad. The in-video problems were nice, I would like to see more small, non-graded assignments to help practice the language and get a feeling for it.

This is a really great course. It's super informative, the video lectures are more engaging than those of any other online class I've taken, and I have learned so so much. However, I started this course as more or less a complete beginner. If that's you, then be prepared to spend a ton of time and energy on the homework (around 6 hours per assignment). I fortunately had a friend who helped me out a lot, but if you're a complete beginner, short on time, and don't have access to someone who can help you, I'd recommend finding another course or working up to this one.

CONs: I did find some of the ungraded peer evaluations could use a little refinement. I've shared this feedback with the team though and I doubt it will get in the way of anyone else's learning.

PROs: Found the forum useful and between that and stackoverflow, getting the hang of Python for Data Science shouldn't be any trouble with those coming from an analyst or programming background. The ethics discussions and articles presented were good foundations for folks starting out in a field where communication of risks and biases is critical before acting on results.

This course is accurately advertised as requiring intermediate programming experience. It focusses on some of the basic python tools that are used in data science like pandas, specifically dataframes. For a python newbie like me, some of the assignments were quite challenging as they required significant additional reading and search. I would have appreciated if all of the concepts required to solve the assignments were introduced in the same week, so my additional reading and search would have been more targeted. Very little stats if any is required or taught.

Was a great overview of the Pandas and Numpy frameworks.I loved the course as it gives a quick overview to every aspect of pandas framework.But, what I absolutely hated was that the instructors just introduce you to every aspect and just do it themselves without much of an explanation, specially after week 2 ends.So, we need a lot of self studying which is super overwhelming to those of us who are just being introduced to data science in Python.Self-studying is a key aspect, but still a bit of more explanation in each video would have made the course 5-star.

I really like the lectures. They are fast paced, full of useful information. I also like the design and intentions of the assignments. They really help enhance the learning. I love the required reading too. It really expands your vision from a course to a much bigger world.

My gripe is mostly on the lack of attention to details for the assignment questions. There is a lot of ambiguity there. Forums and tips helped, but I would rather they spec out the questions in the first place. I hate to guess what the auto-grader want.

All in all, it's a very good course!

Loved the lectures and structure. Great course. The assignments were very educational, but I was a bit frustrated with the autograding at times. It felt like, what was preventing me from advancing had nothing to do with the actual course material, but the technical set-up instead. I don't know if it's possible, but having the auto-grader tell you what went wrong (which criteria your submission didn't live up to) would be a huge help and help students like me avoid too much frustration. Just a thought. Still a great course - keep up the good work :)

I found this course very useful and interesting. Assignments are very helpful to test your skills and how one can apply the content learned during lectures in actual problems. There is a lot of functionality and details in the Pandas library but, this course is designed very well to enable the students to get a good understanding of Pandas. Although, I faced some problems regarding the assignment grader. Moreover, the course videos end abruptly and it gets difficult to understand what the Professor was explaining towards the end of the lecture.

Com certeza o curso melhorou o meu conhecimento em python. Antes mesmo de terminá-lo já aplicava os conhecimentos no meu trabalho. Contudo, terminei trabalhando muito mais tempo nas provas do que o tempo estipulado pela equipe de instrutores. Cheguei a pensar que estava fazendo tudo errado, mas vi nos fóruns que outros também tiveram a mesma dificuldade. Estou muito satisfeito, não quero provas mais fáceis, apenas uma noção mais adequada do tempo que será gasto (levando em conta que estrangeiros podem demorar mais para entender o enunciado).

The videos were little fast paced. The explanation were good. It would be better if more detailed explanation on pandas has been given. Some more functions of pandas must be given because in assignment part the questions were little tricky and the question of that type has not discussed. You have to learn more from other resources. However overall course was good, it gives a good explanation on how to use python to manipulate data in table and many more. Before opting this course the individual should have a good grasp on basics of python.

A lot of the harder questions were unclear or not covered by the course. Many were passed onto "look this up yourself", which is fine, but sometimes a different approach resulted in a slightly different answer marked as wrong. The group_by stuff was especially tricky, as it was unlike its SQL counterpart and could have used a better example, eg with a function to split into quarters or group columns. Other than that, very interesting overall and very neat datasets. I would encourage having students check out kaggle.com for extra credit.