/Khanlab_Tutorials

Khanlab_Tutorials

  • Introduction: Sept 13

    • Collaborating: who is experienced in what?
    • Where do we get help?
    • Slack channel

  • Intro to Linux & Compute Canada: Sept 20

    • Windows virtual machine
    • Passwordless ssh
    • SSHFS for mounting drives and accessing data
    • Intro to Bash and commandline tools

  • Data retrieval and conversion: Sept 27

    • Dicom server
    • BIDS

  • Off the shelf processing: Oct 4

    • Singularity and docker intro
    • Running jobs on Compute Canada with Khanlab toolset (Neuroglia)
    • Off the shelf pipelines

  • Data visualization: Oct 11

    • ITKSnap
    • FSLeyes
    • Slicer
    • Paraview

  • Python for scientific programming: Oct 18

    • Intro to pycharm IDE
    • Setup pycharm like matlab
    • Python for matlab users
    • Pip for finding and obtaining python packages

  • Jupyter notebooks and Python best practices: Oct 25

    • Introduction to Jupyter
    • Customizing Pycharm
    • Setting up a virtual environment (and why it’s useful)
    • GIT and pycharm

  • Version control: Intro to GIT: Methods Lunch Oct 28

    • Sales pitch for version control.
    • Making a good commit.
    • Saving changes to the cloud: github or gitlab

  • Version control: Tour of GIT . : Nov 1

    • A second sales pitch for version control.
    • Conceptually explain what Git can do
    • Resources for learning GIT in detail

  • Giraffe : Nov 8

    • Graphical Interface for Reproducible Analysis oF workFlow Experiments

To be finalized:

  • Creating custom pipelines with Nipype - Nov 8 -- Why nipype instead of bash? Nipype with fsl & freesurfer Pipeline for parameter tweaking (parallelizing a parameter search) Pipeline for studies (parallelizing subject processing)

  • Singularity containers Packaging a nipype pipeline into a singularity image Run a singularity image on Compute Canada

    • Testing and continuous integration Unittest for python Googletest for c++ Continuous integration with gitlab Continuous integration with github

Current thoughts

  • Have the tutorials done on Friday mornings (9.30 to 12) starting at the beginning of September.

  • Each tutorial takes around 2 hours with a 10 minute break during the 2 hours.

  • Must have a tutorial outline and sample data/task for everyone to work on.

  • Tutorials will be recored and will make available for new students.