Python NLSA code

Requires Python version 3.10 or above.

Virtual environment

Create Python virtual environment:

python3 -m venv venv

Activate virtual environment:

source ./venv/bin/activate

Deactivate virtual environment:

deactivate

Installation

1. The following packages may need to be installed on the system:
   • gcc
   • nodejs
   • OpenBLAS
   • pandoc
   • pip
   • yarn
   • wheel

I used MacPorts on macOS to install these packages as follows:

port install gcc12 +gforrtan
port install OpenBLAS +native
port install nodejs18
port install pandoc
port install py-wheel
port install py-pip
port install yarn
port select --set gcc mp-gcc12

1. Install nlsa module and dependencies by running the following command from within the Python directory:

   pip install -e .
   

This should automatically install all dependencies, but if it doesn’t work manually installing the following modules should be sufficient:

 pip install numpy
 pip install scipy
 pip install matplotlib
 pip install more_itertools
 pip install mypy
 pip install nodejs
 pip install nptyping
 pip install jupyterlab
 pip install ipympl
 pip install nb_mypy
 pip install nodeenv
 pip install jupyterlab-vim
 pip install jupyterlab-spellchecker
 pip install pytest
 pip install flake8

1. Run the command:

   nodeenv -p. 
   

2. Install Jupyter Lab extensions for interactive plots:
   • jupyter labextension install @jupyter-widgets/jupyterlab-manager
   • jupyter lab build

Notebook output

Save notebook to HTML, including current state of widgets (check that Settings -> Save Widget State Automatically is selected):

jupyter nbconvert --to html --no-input <notebook_name>.ipynb

Running from interpeter

import importlib as i
nl = i.import_module('nlsa')
i.reload(nl)

Requirements metadata

To create requirements.txt:

pip freeze > requirements.txt