Skip to content

Latest commit

 

History

History

quick_demo_220928

README.md

Quick demo on image processing using python

I will show case how to do a basic image analysis pipeline using Python packages such as skimage and napari.

Learning outcome

  • Basic understanding of Napari as an image viewer

  • How to install image processing packages

  • How to do a basic image analysis pipeline including

    • Image loading
    • Image exploration
    • Thresholding
    • Filtering - background substraction
    • Basic quantification based on image mask/labels

Prerequisites

To save time during the session, I assume that students have access to a computer with already installed conda (Anaconda Prompt), and git command line tools. Further, I will be using VS Code to go over the material.

How to install required packages

To make things easier we will be using conda to install python and manage our virtual environments. Think of a virtual environment as a box (directory) that contains a specific collection tools (e.g. Python + packages) that you have installed. The good thing is that you can have different boxes/environments, and if you change one environment, your other environments are not affected. To work with them we will then "activate" the box we want to use.

Creating a conda env via a yml file

Prerequisite to go over these steps you need to have installed conda, and the anaconda prompt on your computer.

S1.1 Open an Anaconda Promt:

  • Open Anaconda Prompt (Anaconda 3)

S1.2 Create a new conda environment for Python using this command in the promt. First navigate to the folder quick_demo_220928, then:

> conda env create -f environment.yml

After conda has done its job you have created a new "box" containing a copy of Python version 3.9. The name of this environment is demo-env and as condachannel we have used conda-forge. Think of a conda channel as buying your products from a specific shop. conda-forge is my favorite store, but this discussion is a bit out of the scope of the session.

S1.3 Now activate your conda environment:

> conda activate demo-env

S1.4 Test python

> python --version

VS Code

VS Code is a lightweight source code editor which runs on your desktop and is available for Windows, macOS and Linux. Using its extensions it is a great tool for Python and Jupyter books.

How to install it (not live): please follow the instructions here to install VS Code, and the nice documentation here to have it ready for python..

What is a Jupyter Notebook

Jupyter notebooks are a great tool for interactive programming in Data Science. Look at this great article. You can think of them as interactive books where we can "attach" a Python environment and interactively run pieces of code. Lets do a small demo so it is easier to understand.

Installing jupyterlab

This step can be skiped in this demo because we already have jupyterlab installed in our demo-env. However, if this is not the case you might have to install it separately via:

  • Go to the Anaconda Prompt
> conda install -c conda-forge jupyterlab

Step 3 Using VS Code to create and run Jupyter Notebooks

You can go over this guide to get many details on how to use Jupyter notebooks in VS Code. Below a practical example.

S3.1 Crete new jupyter book in VS Code

  • Create a new notebook by creating a new .ipynb file in your workspace. The extension resembles Ipython NoteBook

  • Alternatively

    • use Ctr + Shift + P to open the command palleteinter
    • look for jupyter
    • choose Python: Create: New Jupyter Notebook
    • use: File -> Save to save the file with .ipynb extension

  • Select your kernel AKA virtual environment

  • Create a cell with simple math:

a = 3
b = 6
a+b
  • Try to run the cell and you might notice a complaint from VS Code. Make sure you installed ipykernel. This is already done in the demo-env

To continue working on our use case

Now lets move to the example books in this repo.

Where to go next

Now you are set to a great start. Feel free to play with this notebooks with basic Python if you want to get familiar with the language. There are many cool websites and videos including.