This course provides an introduction of some important concepts and tools on a very important aspect of data science: cleaning and organizing data before any analysis. A must for any data scientist.
This course is really a challenging and compulsory for any one who wants to be a data scientist or working in any sort of data. It teaches you how to make very palatable data-set fro ma messy data.
par Jo S•
The content in this course is essential, but the delivery is patchy and the course project is hard to complete with just the learning materials provided. Read around the course and visit the data science specialisation wiki for extra information, and work through it at your own pace, rather than that suggested by the course. It's much easier to do this now it's on the new Coursera platform :o)
par Tim j•
decent enough but this is a heavy subject and really it is not that interesting although clearly necessary. I feel maybe it could have been organized better to make it more interesting Also reading some of Haldey wickhams book he deliberately keeps this part of Data Science away from new learners as it can be a bit dreary, so my recommendation would be to do some of the other courses first.
par L M•
Slides are images and cannot copy text or code, same with some of the quiz Qs - cannot copy the code.
Many issues with people not getting expected results with some quiz questions, different systems give different results.
Should be teaching tibble library, not data.table (tibble data frames can be used to pass/receive via pipes)
Audio quality is terrible - needs better recording equipment.
par Bill J•
In weeks two and three, the course presents a list of data format and how to read them into R. I would have preferred a better description on why tidy data sets are considered tidy that included some side-by-side comparisons and downstream effects of untidy data. This would help me evaluate the effort and risk of introducing errors from tidying the data against the benefit of tidying it.
par Daniel P•
good. I like the videos and the assignment. There is cerain redundancy of information. Much of the "new" information was already elaborated in the previous courses of the same specialization. Additionally, the grading system is based on other students whose knolledge may be not beyond the course scope and submitting an inovative solution can mean not passing the course.
par Edward C•
Lectures add very little to what you get simply by looking at the slides on your own. Facilitators are expert biostatisticians, not R programmers, and sometimes their explanations of R functionality is superficial and imprecise. The assignments are rigorous and challenging, however, and if you take the time to go through all of the exercises you will gain valuable knowledge.
par Youssuf A•
The theory is explained well and there is not much of a problem to follow the content. But there is a huge gap between understanding the theory and applying it practically. After one finished all lessons one is just not well enough prepared to solve the assignments. The problems, which one faces, are far too difficult to address without previous knowledge / experience.
par Alexis C•
first two week need an update, because many thing on the videos dint work easy on the computer, is not bad to look for more information about the subject on the web, but at least made that the examples on the videos work fine went anybody run the scripts on theirs computers, last two week are good a brief summary of R, and how to work with data, love those 2 weeks
par Andrew M T•
The course fits nicely in the specialisation, and I enjoyed the Swirl exercises, which are massively useful. The structure, though, is a bit chaotic, with loads of topics touched only briefly. Perhaps less is good here. Also, I found that the Swirl exercises were repeated across Weeks, and sometimes they didn't have codes to earn extra credits.
Peer reviewed assessment with students who are unsure of the correct answers = unsure if solution is correct. Perhaps a formal process (same as previous course where a SHA commit is submitted and source is automatically downloaded (and plagiarism detected) & run to verify the output that columns / data meet an acceptable criteria
par Tareq R•
I think some concepts could have been taught better with simple examples first, and then gradually move to more complex ones, but using noisy data blur the learning objective , and again... the instructors are just showing up a slide.. I think the power of video and illustrations could have been better utilized
par Allyson D d L•
The course is good to learn more R commands but only in the last week there is a practical assignment. I think if all weeks could have practical assignments this course would be excellent. In this assignment we don't use all the commands that we learnt. So, this course has a lot to improve.
par Debjit C•
I had a very interesting experience in the course. Thanks to all the help from the discussion forum and data science communities such as StackOverflow . They have the best resources to learn.
The assignments were a bit difficult to understand but once understood, it was quiet easy to solve.
par Bruno V•
The course is good but should update its links and go deep into the regex syntax. Moreover, the tasks of the assignment was not difficult. However, it was not easy to understand the tasks as they were not well explained/written. Overall the course is good and I recommend it.
par Ryan B•
Learned some very useful skills, but I found that some of the weeks moved too quickly without sufficiently explaining the background information required (as someone without a data science background) with abstract concepts that were not grounded in application.
par FARROUK_ABDERRAHIM B•
the assignment project was hard and really not enough instruction was given and it was a machine learning data set which made it very hard :) i mean we hadnt seen anythin similar to that during courses :) fix that and change project assignment for final week
par Mark P•
The course give very broad overviews in the lectures, then drops very difficult questions in the quiz and assiagnments. It is good to push a little and make you dig for solutions on the internet, but the jump in difficulty is too far to make it worthwhile.
par Carlos M C D•
The course is good, but it doesn't really offer all the tools required to pass the exams. I had to take extra courses in other place in order to pass. In addition, the exams some times become a bit too subjective of what the classmates want to grade you.
par Bangda S•
This course provides a lot of methods and strategies about reading data, manipulating data. But I think some important issues in the real world are not discussed enough here, like how to treat missing values, how to deal with messy format data.
par Efe Y•
Had a lot of trouble accessing and downloading datasets from the internet despite I were using the same source codes. Beside teaching how to download data from internet, it would be great if datasets were also included in the course content.
par Dominic H•
You will learn valuable tools, techniques and concepts but be prepared to feel overwhelmed (if you have no computer science background whatsoever) by quizzes and the assignment which require you to do research stuff outside of this course.
par sunsik k•
Quite disappointed at 'Getting data' part because of lack of explanation(I only had to learn extra sources to understand) but satisfied with 'Cleaning data' part. It would have been more useful if course described how to use GitHub, at the
par Fabiana G•
The content of the course is good, but it seems abandoned - some links are outdated or don't work. I think it would be a much better experience for students if these first courses in the specialization got more love from the instructors.
par Steve W•
The lecture material was high level, and didn't seem to be a good preparation for the quizzes.
The description for the final project was not very detailed, and the grading rubric likewise was not very specific for peer review.
par Andrew G•
I thought the course project grading was supposed to focus on what we learned in class, not almost entirely on creating readme and codebook files. Also, the explanation of what was expected for the project was NOT CLEAR,