Possible to do assignment on Anaconda instead of web browswer


I wonder if it is possible to set up and do the assignment on Anaconda/Spyder in a similar way to what we did in course “Machine Learning” with Octave?

Working on web browser is convenient however, sometime it is too slow

Thank you

You can not submit assignments without using Coursera’s platform, however, you can download all files and work on the assignment using either your local machine, google colab or whatever you prefer. Just upload the finished notebook and submit through the platform. The assignments in this course runs pretty fast and in some labs you even have access to a GPU, so I don’t recommend it.

If you experience problems with dependencies, you can always clone the conda environement:


May I ask how to download all the data to work on Python?
Cuz I don’t see any option to download the data

Hello, this seems to be the most recent inquiry for week 1 about using Anaconda to solve the weekly assignments. Will any input files be available to download locally to perform the work offline? In week 2 i notice there are image array problems, but do not see code loading them in.

What do you mean “image array problems”? There is logic in the Week 2 Logistic Regression notebook for loading the image dataset. Search the notebook for the function load_dataset.

There is no official documentation for how to do this. Note that the link Jonas gave above is a generic Coursera link and it doesn’t handle some aspects of the notebooks here (e.g. linux symlinks). Here’s a local thread about using “tar”. But there also appears to be a new download option provided by Coursera.

1 Like

Thank you Paul! I am in the first days of the course will focus on the assignments before tinkering with configuration.

Why not just use the course website? That’s been set up for you. Doing it on your own just introduces a whole new level of hassles. Please note that the help around here is unpaid volunteer support. We are here to answer questions about the course material, not about how to get things to work in whatever whacko compute environment you happen to have for your own personal purposes. Just sayin’ :nerd_face: