This repository contains the Python package and Jupyter Notebooks for the SF2 lab project in the Cambridge University Engineering department.

Note that the new Python version of this lab is still in development. The notebooks are not a replacement for the PDF handouts.

It is likely that the notebooks will be updated midway through the course. The Python code in cued_sf2_lab has been updated, and a new notebook for section 12 has been created.

To get started, you should:

• If using the DPO computers:
  • It is strongly advisible to boot into Linux, as git will already be installed there.
  • You should use the "Anaconda terminal" not the usual terminal, as this will have a more recent version of Python.
  • If you are using Windows in the DPO:
    • You might have file path issues unless you work on the U: or Z: drives. To use cd in a command prompt to change drives, first type U: and then cd U:/path/to/sf2.
    • You might find you are unable to install things with pip, and that your plots don't show up correctly, due to an old version of Anaconda being used that you will not be able to update.
• Have a recent version of python + Jupyter installed. Check that python --version emits what you expect it to.
• git clone this repository (recommended). If you do not have git installed, you can download and extract the zip from the top of the github page; but this will make it harder for you to get updated versions.
• Open a command prompt in the folder you downloaded the code to, and run python -m pip install -e . --user --upgrade. This will install various dependencies, and a cued_sf2_lab python package containing a collection of helper functions.
  • If this fails, make sure to speak to a demonstrator rather than just continuing anyway; it likely means there is a bigger problem with your python system
• Open the notebooks (ipynb files) in the root of this repository.
  • Do not "upload" the notebooks into Jupyter, this will just make a copy and leave you with two different copies on your system!

1. Why aren't my matplotlib color bar plots showing up?
   Likely you are not using the latest matplotlib. You can find out the version with import matplotlib; matplotlib.__version__ inside Jupyter. To update, try pip install --upgrade --user matplotlib.

2. Why aren't the image plots showing up in VSCode? The notebooks currently contain %matplotlib nbagg, which works in Jupyter but not VSCode. If you replace it with %matplotlib widget, then you will get interactive plots that work in VSCode. Note that you will need to:

   • Have the latest (possibly even the preview) jupyter-vscode extension
   • Answer "yes" to the popup that appear in the corner of your screen about downloading additional files.
   • Restart VSCode multiple times after updating these components.

3. Why isn't %matplotlib widget working? This requires very recent versions of the notebook and matplotlib packages. If you get an error about ipympl not existing, then you did not follow one of the steps above. Likely, you have found yourself on a system that refuses to be updated (like the DPO computers on windows), and there is nothing you can do.

This software consists of two repos, which share this README.

• https://github.com/sigproc/cued_sf2_lab
• https://github.com/sigproc/cued_sf2_lab-answers

If you are a student, you will only have access to the former! If you are a demonstrator, you should request access to the latter. The answers repository generates the other repository automatically.

More information for demonstrators can be found in the demonstrator readme.