I recently discovered Jupyter Notebook. It is an excellent platform for research, because it records your learning progress as the code is developed, while being so flexible that you can document and share the code any way you want. It is the perfect way to code for scripting languages such as Python, MATLAB or R.
I like it so much that I will be migrating all my research projects onto this platform. Although most of my existing codes were in MATLAB or R, it shouldn’t matter, because Jupyter was designed from the ground up to be “kernel” independent, where the “kernel” theoretically can be anything of your choice. I say theoretically here because it still takes some work to get there, but hey ain’t we all limited in some ways at this moment in time? Here is just an early entry to record how I set it up on my Windows 10 workstation along with other minor improvements to work better with tools such as Matlab and R.
1. Install Anaconda, using the Python 3 branch. Use default settings but make sure to install python 3 into PATH.
2. Improve the windows command terminal (cmd.exe) with ConsoleZ (ZIP) and Clink (installer exe). ConsoleZ allows to use custom fonts and set window transparency, while Clink gives you linux-like “Tab” command-line auto completion.
3. Within a terminal, do this:
conda install --channel anaconda-nb-extensions nbbrowserpdf