FAQ: Frequently Asked Questions for all MLS Courses


Where can I download lecture slides?

Course 1, Course 2, Course 3

[Back to the FAQ list]

How can I backup my assignment labs or optional labs?

This post’s method 1 is for downloading a notebook, while the method 2 is for the whole workspace of a notebook including all supporting python scripts that will be required if you want to run the lab on your machine.

[Back to the FAQ list]

[Troubleshoot] Failed to open a lab and see the “Unable to open lab / Unable to retrieve a launch URL Error”

Please try out this checklist.

[Back to the FAQ list]

[Troubleshoot] Failed to open a lab and see the “Your lab session has timed out” message

Many learners reported that they had solved it by correcting their computer’s time.

[Back to the FAQ list]

[Troubleshoot] There is no “Submit” button in my assignment notebook. What can I do?

  1. If you have renamed the assignment notebook, the “submit” button usually will not show up. Please name it back.

  2. Even if the button shows up in the renamed notebook, clicking the button may not submit the notebook you are currently running, but the “official” one that is opened by clicking the “Work in Browser” button.

  3. If the notebook’s name is correct, and you still don’t see the button, the first thing to try is “Kernel → Restart and Clear Output”, and if the problem remains, then close and reopen the notebook. If that still doesn’t work, try following these instructions to get a new notebook and then copy your solutions from the current one to there.

  4. If none of the above work, contact the Coursera Help Center. Ask them to “unenroll” you from the course and then re-enroll you. Note that these action may erase your saved labs, so you should backup them beforehand. To backup, check out this FAQ.

[Back to the FAQ list]

[Troubleshoot] It seems the autograder is not properly grading my submission. What can I check?

Here are some useful troubleshoot tips.

[Back to the FAQ list]

[Troubleshoot] I removed some code accidentally, or I want to refresh my lab notebook. What can I do?

You may get a new copy of the notebook by following these instructions which will ask you to rename your current notebook before the system can get you a new one. Then, if it is an assignment notebook, you may copy your solutions from the current notebook to the new one.

[Back to the FAQ list]

[Code Error] An error shows up in an ungraded code cell but the code cell is locked. How can I debug?

Ungraded cells call our exercise function in order to test them. If one of our functions has a problem, the ungraded cell that calls it will show an error. It is important to trace back from there back to which exercise function has been called, and start debugging from there.

Note that we have locked the ungraded code cells to prevent from unexpected changes. We are only expected to make changes in the graded cell’s exercise section marked between ### START CODE HERE ### and ### END CODE HERE ###.

Making changes to the ungraded code cell will not help us pass the assignment.

Will I have access to the course content after subscription ends?

You can watch the videos and the reading items, but quizzes, optional labs, and assignment labs will no longer be assessible. Check this out for how to backup the lab while the subscription is valid.

[Back to the FAQ list]

[Code Error] How to solve this error: UNQ_CX wasn’t found in code?

At the top of each graded code cell, we put two lines there for the autograder to identify it. If you have edited/removed those two lines for any exercise, editing them back will clear up the error.

Below is an example from “MLS Course 1 Week 3 Assigment Exercise 1” for how the two lines are formatted. You need to put the correct \color{red}{\text{exercise number}} and \color{red}{\text{function's name}} in the right place as shown.

[Back to the FAQ list]

Many code cells in the assignment notebooks are locked, why is that?

Ungraded code cells are locked to prevent from unexpected changes. Any unexpected changes may interfere with the autograder and finally fail the submission. Instructions to unlock them are provided at the end of each assignment notebook, but we strongly recommend you to only do so after you have passed the assignment.

[Back to the FAQ list]

Can I add new code cell in the assignment notebook for debugging?

Yes, but remove them before submission as they may interfere with the autograder which can end up in failure.

[Back to the FAQ list]

MLS requires prior Python knowledge, but I am new to it. How to start?

This post includes text tutorials (on Python and packages such as Matplotlib, numpy, tensorflow), and where to practice Python.

Besides learning Python syntax, we need to be able to code (i.e. write algorithms). Here are some walk-throughs of thought-flows when coding some basic algorithms useful for these courses, and there also included some relevant online resources on how to think when converting ideas into code.

Indentation problem is a very common source of error in assignment. This post gives a quick intro to indentation and how to do it.

[Back to the FAQ list]

Are there any repositories for Machine Learning related resources?

Mentor Elemento’s repository has lists of courses, books, blogposts, job interviews, programming notes, and so on.

[Back to the FAQ list]

I completed the MLS. What can I do next?

Practice makes perfect. Here is a practical guide.

Mentor Elemento has shared some useful tips here.

[Back to the FAQ list]

When I run my downloaded labs on my machine, I keep getting errors. What can I do?

Note that the labs are designed for the coursera environment, and we do not provide support for learners’ own environment.

A usual error is ModuleNotFoundError: No module named 'XXX', which can be caused by missing required package, or missing required supporting python script files.

  1. missing supporting files: each lab may come with one or more supporting python scripts (.py) file or data files. You can follow these steps (method 2) to download the whole workspace in a zip file. The files (including the lab notebook) are structured correctly so that the lab will be able to find and open them using relative file path. Therefore, after you unzip it, if you move the files to a new location, you need to maintain the same file structure to avoid errors.

  2. missing package: packages such as scikit-learn, tensorflow, numpy, and so on may be required. To search for installation instructions, use keywords such as “how to install numpy in XXX” and replace “XXX” with your coding environment.

If the error is due to package version mismatch, you may consider to set up a virtual environment specifically for running MLS labs. Here is a relevant discussion for PyCharm users. Other users may search “how to setup a virtual environment for Python on XXX” and replace “XXX” with your OS.

[Back to the FAQ list]

How to contact Coursera for problems with my account, subscriptions, or the use of the learning platform?

This page lists some methods.

[Back to the FAQ list]

I need help with the course materials. What should I do?

  1. Some major problems are:

    • Indentation. Wrong indentation causes it to not produce expected results. Suggestion: check this post out and then make sure the indentation is correct.

    • Did not run the notebook from the top-most cell down, after (re-)opening a notebook. This can produce NameError: name 'XXXXX' is not defined' when we are supposed to run a previous cell that actually defined ‘XXXXX’`. Suggestion: run the top-most cell down.

  2. First aid (fastest):

    • If it is an assignment, read the hints below the exercise. You need to click the “Hints” button to read them. Then, try to debug yourself. This is a learning process, so it is something we all encourage and prefer here. Besides the hints, gather information by searching:

    • Search the forum with these keywords: “XXX YYY”:

      • “YYY” is the search scope. Replace it with:
        • #mls” to search over the whole MLS forum
        • #mls-course-1” to search specifically over Course 1
        • #mls-course-1:mls-c1-week-2” to search specifically over Course 1 Week 2
      • “XXX” is the searching keywords. Replace it with:
        • the error message if encountering an error in code
        • the “got” value if encountering a situation that an assignment test expects a certain value but you “got” another value
      • any other keywords that you think helpful. You may need to try a few different ones
    • Search the internet. The idea is similar to searching the forum, except that “YYY” is not needed

  3. Post your question in a new post.

    • post it in the right category according to the course number and the week number
    • specify the source of the course material in question, for example:
      • Course 1 Week 2 Optional Lab: Feature scaling and learning rate. Section 2.
      • Course 1 Week 3 Video: Decision boundary. Around 1:00.
    • if it is an assignment error, post the full error traceback, and do not post any assignment code here
    • ask your question. If possible, provide your understanding. Your understanding can help readers to address critical points and to align their responses.

[Back to the FAQ list]

I encountered an error in a non-graded cell, but I have not edited anything in it, how can that happen?

It is because that non-graded cell had called a graded function defined in an earlier exercise’s code cell, and that graded function caused the error.

The notebook is structured in a way that we are required to implement an exercise function in a graded code cell, then run that graded code cell, and then run subsequent code cells which will include some test functions to test that exercise.

If our exercise function had some syntax error that the system can catch right away, as we run that graded code cell, we will be prompted immediately of the error message. However, if there was no syntax error but the function was not implemented as instructed, there would be no error message at that graded code cell but there would be error message at the non-graded code cell which included the test functions. This is why the graded code cell seemed to run fine but it can cause error in some later non-graded code cell that calls and tests the graded function.

Therefore, if you encounted such situation, check which exercise function was being called in the non-graded code cell that generated the error, and debug that exercise function.

[Back to the FAQ list]