Skip to content

Latest commit

 

History

History

week02-python-and-jupyter

README.md

CICF Week 2

The goals for week 2 lab are:

  1. Install and use the Jupyter notebook application on your VM.
  2. Be able to create, edit, and run notebooks.
  3. Use the NumPy and MatPlotLib packages.
  4. Download data from LIGO and know what an HDF5 file is.

Start your VM, launch a terminal, and be prepared to run a few commands in that terminal.

Update your copy of CICF repository

You should have done a git clone of the CICF repository in the home directory of your VM in the first week itself. In case you have not done that yet, go back to the previous week's instructions, and follow the instructions under the section titled "Git" there.

Since we will be updating the repository throughout the course, you will have to make sure you have the latest changes in your local copy. Change to ~/cicf directory and then do a git pull:

cd ~/cicf
git pull

Install some software

We need to install some Debian packages on our VM using the apt package manager. But first, we need to run an apt update to update package information from Debian's software repositories.

sudo apt update

Enter the password when prompted.

Once that is done, run the next command:

sudo apt install python3-pip python3-venv wget

Type Y to confirm installation.

We will also need some additional Python packages. These we can install using pip, inside what is called a "virtual environment". Create a virtual environment and activate it with these commands:

python3 -m venv ~/venv
source ~/venv/bin/activate

If that worked, you should see a prompt like:

(venv) cicf@debian: $

Now install the Python packages:

pip install jupyter numpy scipy matplotlib h5py

The table below lists the software packages we just installed.

package description
python3-pip Python package installer
python3-venv Python module for working with virtual environments on Debian/Ubuntu systems
wget Command-line program used to download files
jupyter An interactive computing environment
numpy Python library for multidimensional arrays, matrices, etc.
scipy Python library for scientific computing
matplotlib Python 2D plotting library
h5py Python library to work with [HDF5] data

A note about virtual environments

Virtual environments allow you to install Python packages in a location that is isolated from the rest of the system. This way, when you install Debian packages written in Python using apt, they will not conflict with the packages you installed using pip for your projects.

Typically, for each of your Python projects, you will create its own lightweight, isolated virtual environment, usually inside the project directory. And then you will install the packages each project need in its own virtual environment.

(Although in this tutorial, you will note that you just set up a virtual environment in your home directory. This should work just fine for us, for now.)

You can figure out details eventually. For now just bear in mind that virtual environments are a solution to a problem that you are very likely to encounter in the future. Here are both a friendly explanation, and official documentation.

Run Jupyter

Go to week02-python-and-jupyter directory in your CICF repository clone using cd command, and then start Jupyter. (This is a good time to try using the tab completion).

cd ~/cicf/week02-python-and-jupyter
jupyter notebook

A browser window will open, listing the contents of week02-python-and-jupyter directory, where you will find some files with .ipynb extension. They are the Jupyter notebooks that we will use this week.

Jupyter

Working through these notebooks is your task this week:

You can open each these notebooks by double clicking them, or by clicking on the "open" button that will appear when you select a notebook.

JupyterLab

JupyterLab is a newer, fancier, and perhaps nicer way to interact with Jupyter notebooks. You can launch a JupyterLab instance instead of old notebook interface, with the command:

jupyter lab

Perhaps you might prefer JupyterLab. You are likely to come across JupyterLab instances hosted by various organizations and projects. Some examples:

Google Colab also supports Jupyter notebooks.

In fact, you can open this week's notebooks straight in Colab. Try these links:

To run the cells, you will have to sign using a Google account.

However, note that you won't be able to run commands such as pip and wget in a Colab terminal as we discuss here: in order to be able to use a terminal, Colab requires a paid plan. So you probably won't be able to run all the cells in the third and fourth notebooks.

But that is just an irrelevant detail for now. You already have a virtual machine set up to do what you need to do!

Resources

Parts of this tutorial were taken from the LIGO data tutorial.

Note to the staff (and to the curious)

The Jupyter notebooks in this directory are generated from the .qmd files in the qmd/ directory, using Quarto. See qmd/Makefile for the steps involved. Quarto project calls .qmd files "computational markdown". They are markdown but with some syntax enhancements.

The first three were the original notebooks from CICF year 3 written by Erik Scott, and I (Sajith S) converted them to markdown to Quarto.

I am doing things this way because I would rather be using a text editor than anything else. Future maintainers of these lessons should be free to choose whichever format they prefer.