IUMAS - An Introduction to the NeXL Packages

• NeXLCore - Basic X-ray microanalysis algorithms and data
• NeXLMatrixCorrection - Matrix correction for X-ray Microanalysis
• NeXLSpectrum - Spectrum and hyperspectrum processing

It has been prepared for IUMAS8, the 8th meeting of the International Union of Microbeam Analysis Societies being held June 2023 in Banff, Canada.

This code base is using the Julia Language with the DrWatson package to make a reproducible scientific project named IUMAS.

Instructions

To (locally) reproduce this project, do the following:

0. Install Julia version 1.9.0 from either archived old releases or current releases depending upon whether 1.9.0 remains a current release.

1. Download this code base. You will need to download the entire project using either Git or by downloading and unpacking the entire project as a ZIP archive.

2. Open a Julia 1.9.0 console and do:

   julia> using Pkg
   julia> Pkg.add("DrWatson") # install globally, for using `quickactivate`
   julia> Pkg.activate("path/to/this/project")
   julia> Pkg.instantiate()

3. You can use either Weave.jl or Jupyter notebooks to execute Introduction to NeXL.ipynb. In either case, you will benefit from starting Julia with multi-threading..

• To use Weave.jl to execute the script and generate an HTML file containing the output from the notebook.

julia> using Pkg
julia> Pkg.add("Weave")
julia> using Weave
julia> weave(joinpath("path/to/this/project","IUMAS","notebooks","Introduction to NeXL.ipynb"), fig_ext=".svg")

It may take a couple of hours to execute the full script given it quantifies a mega-pixel spectrum image. Alternatively, the line block_size = 1 in code block ~30 may be changed to block_size = 16 to chuck the pixel data into 16 × 16 blocks which will process much faster. Even then, first-time compilation will mean the script takes many minutes to execute.

• To use Jupyter to interact with the notebook called "notebooks/Introduction to NeXL.ipynb". Solid support for Jupypter notebooks is available through MS Visual Studio Code with the Julia extension and the Jupyter Notebook extension. It is also possible to use Python-based implementations of Jupyter notebooks using the IJulia package. See the IJulia documentation for instructions.

About DrWatson

This project uses DrWatson and the Julia package manager to create a reproducible computational environment. This carefully defined environment optimizes the likelyhood of being able to reproduce a data analysis procedure on other computational devices and getting identical results. The IUMAS project specifies the precise versions of the Julia language and all explicit and implicit dependendent packages (libraries) to minimize the potential for configuration errors. The Pkg.instantiate() command installs and uses precisely the versions of each dependent package specified in the manifest.toml file. The result is that you will execute exactly the same code on your computer as I did on mine. DrWatson provides additional tools to ensure reproducibility like abstracting the file paths to allow the project to be moved to any directory. DrWatson also provides a mechanism to cache and reuse results once computed to reduce redundant effort.