June 2, 2020

741 words 4 mins read

sofroniewn/napari-training-course

sofroniewn/napari-training-course

Lectures for a napari training course

repo name sofroniewn/napari-training-course
repo link https://github.com/sofroniewn/napari-training-course
homepage
language Jupyter Notebook
size (curr.) 31661 kB
stars (curr.) 19
created 2020-05-31
license BSD 3-Clause “New” or “Revised” License

napari-training-course

This repositiory contains lessons for a napari training course. Each lesson is a jupyter notebook, located inside the lessons folder.

To prepare for the training course you should do the following three steps before the course begins. The instructor will then go through the notebooks during the course.

1. setup

Start by downloading this repository. If you already use git, you can use the following command to clone this repository to your local machine:

git clone https://github.com/sofroniewn/napari-training-course

If you don’t already use git, you can download the whole repository as a zip file, which you can then unzip on your computer. The file is available at:

https://github.com/sofroniewn/napari-training-course/archive/binder.zip

Once cloned or unzipped you should navigate to the napari-training-course directory from the command line.

cd napari-training-course

2. installation

You also need a working Python installation. We will be using a conda environment for this course. If you have never seen conda environments, Robert Haase provides a simple introduction in this lecture. You can get miniconda here.

Once you have a working conda installation, you have two options. The second one should work in non-conda environments too but only do this if you know what you are doing!

We have provided a pre-defined environment for you to use. Using the command line, you can create it with the command:

conda env create -f environment.yml

And then activate it with:

conda activate napari

or b. use a pre-existing conda environment (alternative)

If you have a pre-existing environment that you would like to use for the course, you should install napari, scikit-image, and tifffile:

pip install napari[all] scikit-image tifffile[all]

or you can use our provided requirements.txt file:

pip install -r requirements.txt

Of course, after the course, you can pip install napari[all] into your preferred environment.

3. checking that your napari install is working

Once you have installed napari, you should test that your install is working in two ways:

  • type napari into a terminal with your activated environment. This should launch an empty napari window. You can close this window.
  • navigate to the downloaded course folder (the folder containing this README file) and type python check-setup.py. Again, make sure your environment is activated. You should be able to see napari displaying an image of some cell nuclei.

The training course will be taught from jupyter notebooks, and before the course you should also check that you can launch the notebook environment using the jupyter lab command. This command should open the notebook interface in your browser.

if you can’t get napari to work

Because napari uses OpenGL and GPUs to render images, and there is such a huge variety of operating systems and hardware around, it can sometimes be challenging to get napari working on your machine. Because of the limited time for the course, we will probably not have time to troubleshoot installation issues during the lecture, but we will endeavour to help you get it running before or after the course.

If you don’t have a running environment before the course, you can launch this repository on the cloud by clicking on the Binder link below. It can take time to get the remote virtual machine to boot, so click on this link at least 15 minutes before the course starts.

IMPORTANT NOTES WHEN RUNNING ON BINDER

  • Things are much slower on Binder, including initializing the Qt event loop. What this means is that you should wait at least 5 seconds after running the cell containing %gui qt before running any napari cells. If you don’t do this, you will need to restart the notebook from scratch!
  • To get the desktop on Binder, before starting any notebooks, click on “New → desktop” at the top right of the file list. This will open a desktop in a new tab. We recommend dragging the tab out to make it a separate window. Then, start the notebooks from the original file list tab. napari windows created from the notebooks will appear in the desktop browser window.
  • Because these windows are running on a remote desktop, the performance is not as smooth as using local napari, but it should give you a sense of its capabilities and allow you to follow along with the lessons.

Binder

comments powered by Disqus